CN107181948A - Method for encoding images, picture coding device and image encoding/decoding device - Google Patents

Method for encoding images, picture coding device and image encoding/decoding device Download PDF

Info

Publication number
CN107181948A
CN107181948A CN201710263791.0A CN201710263791A CN107181948A CN 107181948 A CN107181948 A CN 107181948A CN 201710263791 A CN201710263791 A CN 201710263791A CN 107181948 A CN107181948 A CN 107181948A
Authority
CN
China
Prior art keywords
mentioned
context
block
control parameter
value
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Granted
Application number
CN201710263791.0A
Other languages
Chinese (zh)
Other versions
CN107181948B (en
Inventor
笹井寿郎
西孝启
柴原阳司
杉尾敏康
谷川京子
松延彻
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Sun Patent Trust Inc
Original Assignee
Sun Patent Trust Inc
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Sun Patent Trust Inc filed Critical Sun Patent Trust Inc
Publication of CN107181948A publication Critical patent/CN107181948A/en
Application granted granted Critical
Publication of CN107181948B publication Critical patent/CN107181948B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N19/00—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
    • H04N19/10—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding
    • H04N19/102—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding characterised by the element, parameter or selection affected or controlled by the adaptive coding
    • H04N19/13—Adaptive entropy coding, e.g. adaptive variable length coding [AVLC] or context adaptive binary arithmetic coding [CABAC]
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N19/00—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
    • H04N19/70—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals characterised by syntax aspects related to video coding, e.g. related to compression standards
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N19/00—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
    • H04N19/10—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding
    • H04N19/102—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding characterised by the element, parameter or selection affected or controlled by the adaptive coding
    • H04N19/103—Selection of coding mode or of prediction mode
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N19/00—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
    • H04N19/10—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding
    • H04N19/102—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding characterised by the element, parameter or selection affected or controlled by the adaptive coding
    • H04N19/103—Selection of coding mode or of prediction mode
    • H04N19/105—Selection of the reference unit for prediction within a chosen coding or prediction mode, e.g. adaptive choice of position and number of pixels used for prediction
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N19/00—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
    • H04N19/10—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding
    • H04N19/134—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding characterised by the element, parameter or criterion affecting or controlling the adaptive coding
    • H04N19/136—Incoming video signal characteristics or properties
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N19/00—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
    • H04N19/10—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding
    • H04N19/134—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding characterised by the element, parameter or criterion affecting or controlling the adaptive coding
    • H04N19/136—Incoming video signal characteristics or properties
    • H04N19/137—Motion inside a coding unit, e.g. average field, frame or block difference
    • H04N19/139—Analysis of motion vectors, e.g. their magnitude, direction, variance or reliability
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N19/00—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
    • H04N19/10—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding
    • H04N19/169—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding characterised by the coding unit, i.e. the structural portion or semantic portion of the video signal being the object or the subject of the adaptive coding
    • H04N19/17—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding characterised by the coding unit, i.e. the structural portion or semantic portion of the video signal being the object or the subject of the adaptive coding the unit being an image region, e.g. an object
    • H04N19/176—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding characterised by the coding unit, i.e. the structural portion or semantic portion of the video signal being the object or the subject of the adaptive coding the unit being an image region, e.g. an object the region being a block, e.g. a macroblock
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N19/00—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
    • H04N19/10—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding
    • H04N19/189—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding characterised by the adaptation method, adaptation tool or adaptation type used for the adaptive coding
    • H04N19/196—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding characterised by the adaptation method, adaptation tool or adaptation type used for the adaptive coding being specially adapted for the computation of encoding parameters, e.g. by averaging previously computed encoding parameters
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N19/00—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
    • H04N19/50—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using predictive coding
    • H04N19/503—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using predictive coding involving temporal prediction
    • H04N19/51—Motion estimation or motion compensation
    • H04N19/583—Motion compensation with overlapping blocks
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N19/00—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
    • H04N19/90—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using coding techniques not provided for in groups H04N19/10-H04N19/85, e.g. fractals
    • H04N19/91—Entropy coding, e.g. variable length coding [VLC] or arithmetic coding

Abstract

The picture decoding method of the mode of the present invention includes context rate-determining steps, and the context used process object block is determined in multiple contexts；In context rate-determining steps, the signal kinds of the control parameter of process object block are judged；In the case where signal kinds are the 1st species, using using left piece and the 1st condition of upper piece encoded control parameter abutted with process object block, context is determined；In the case where signal kinds are with the 1st diverse 2nd species, using it is unfavorable with piece and left piece encoded control parameter and utilize fixed value set in advance the 2nd condition, determine context；1st species is to represent to deal with objects whether block is divided into multiple pieces of mark or indicates whether that the mark that block is skipped will be dealt with objects；2nd species is the mark for representing to deal with objects the coefficient whether the motion vector composition of block includes brightness and aberration with the mark of the difference of its predicted value and expression process object block.

Description

Method for encoding images, picture coding device and image encoding/decoding device
The present invention is the Chinese Patent Application No. submitted for 22nd for 06 month in 2012 of the applicant for 201280025858.2, Entitled " picture decoding method, method for encoding images, picture decoding apparatus, picture coding device and image encoding/decoding The divisional application of device ".
Technical field
The present invention relates to method for encoding images, picture coding device and image encoding/decoding device, more particularly to using calculation Method for encoding images, picture coding device and the image encoding/decoding device of art coding or arithmetic decoding.
Background technology
The picture signal of natural image has statistical fluctuation, and the fluctuation is in the characteristic of mobility.Using statistically One of the entropy code mode of mobility of fluctuation in, have Context-Based Adaptive Binary Arithmetic Coding (CABAC, the adaptive binary arithmetic coding based on context) (with reference to non-patent literature 1).The CABAC modes exist H.264/AVC, ITU-T/ISOIEC standard for video coding, are used in mode.
Hereinafter, the meaning of term used in CABAC modes is illustrated.
(1) " Context-Based Adaptive " (based on the adaptive of context), is to make encoding and decoding method certainly It is adapted to the variation of statistical fluctuation.That is, " Context-Based Adaptive " could also say that encodes by some code element Or in the case of decoding, be matched with the condition of surrounding generation event predicted as the probability of happening of the code element it is appropriate general Rate.For example, in coding, in the case where determining some code element S probability of happening p (x) of each value, using to actually occur The row F (z) of event or event is the probability of happening of the SNNP of condition.
(2) " Binary " (binary system) refers to show code element with binary array.On the code element represented with multivalue, one The secondary two-value arrangement for being deformed into referred to as bin string.And for each arrangement key element, while switching uses prediction probability (bar Part probability), while by the event that there occurs two values which carry out bit sequence.Thereby, it is possible to than signal kinds The more detailed unit of unit (binary number wants primitive unit cell) is managed (initialization and renewal) to the probability of value (with reference to non-patent Fig. 2 of document 1 etc.).
(3) it is so-called " arithmetic " (arithmetic), refer to above-mentioned bit sequence generation processing be not by the correspondence of table, But exported by calculating.Using H.263, MPEG-4 and H.264 in variable length code table coded system in, even Code element with the probability of happening probability of happening higher than 0.5 (50%), it is also desirable to which each value for enabling code element to take corresponds to 1 Binary system ordered series of numbers (bit sequence).Thus, it is also most short to make 1 bit and 1 code element foundation pair on probability highest value Should.In contrast, arithmetic coding can show the generation under the higher probability of event with below the bit of integer value 1.For example, There is probability of happening of the value of the 1st binary number for " 0 " more than signal kinds as 0.9 (90%), the as event the 1st In the case that the continuous n times of value of individual binary number are " 0 ", it is no longer necessary to each value " 0 " by the data output n times of 1 bit.
Prior art literature
Non-patent literature
Non-patent literature 1：Detlev Marpe,et.al.,"Context-Based Adaptive Binary Arithmetic Coding in the H.264/AVC Video Compression Standard",IEEE Transaction on circuits and systems for video technology,Vol.13,No.7,July 2003.
Non-patent literature 2：Joint Collaborative Team on Video Coding(JCT-VC)of ITU-T SG16WP3and ISO/IEC JTC1/SC29/WG11 5th Meeting:Geneva,CH,16-23March,2011JCTVC- E603,ver.7”WD3:Working Draft 3of High-Efficiency Video Coding”http:// phenix.int-evry.fr/jct/doc_end_user/documents/5_Geneva/wg11/JCTVC-E 603- v7.zip
Non-patent literature 3：Joint Collaborative Team on Video Coding(JCT-VC)of ITU-T SG16WP3and ISO/IEC JTC1/SC29/WG11 4th Meeting:Daegu,KR,20-28January,2011," Common test conditions and software reference configurations",JCTVC-E700
Non-patent literature 4：Gisle Bjontegaard,“Improvements of the BD-PSNR model," ITU-T SG16Q.6Document,VCEG-AI11,Berlin,July 2008
Brief summary of the invention
Problems to be solved by the invention
But, in such method for encoding images and picture decoding method, it is desirable to (the storage used of memory usage amount Device capacity) reduction.
The content of the invention
So, it is an object of the invention to provide a kind of method for encoding images that can cut down memory usage amount or image solution Code method.
The means used to solve the problem
In order to achieve the above object, the method for encoding images of a technical scheme for the present invention, using arithmetic coding, its It is characterised by,
Including：
Binarization step, passes through the control parameter binaryzation by block is dealt with objects, generation two-value arrangement；
Context rate-determining steps, determine the context used above-mentioned process object block in multiple contexts；And
Arithmetic coding step, arithmetic coding is carried out by using the above-mentioned context determined to the arrangement of above-mentioned two-value, raw Into bit sequence；
In above-mentioned context rate-determining steps,
Judge the signal kinds of the control parameter of above-mentioned process object block；
In the case where above-mentioned signal kinds are the 1st species, using using with above-mentioned process object block abut left piece and 1st condition of upper piece encoded control parameter, determines above-mentioned context；
In the case where above-mentioned signal kinds are with the above-mentioned 1st diverse 2nd species, using it is unfavorable with it is above-mentioned upper piece And above-mentioned left piece encoded control parameter and utilize the 2nd condition of fixed value set in advance, determine above-mentioned context；
Above-mentioned 1st species is to represent whether above-mentioned process object block is divided into multiple pieces of mark or indicate whether will be upper State the mark that process object block is skipped；
Above-mentioned 2nd species be represent the motion vector composition of above-mentioned process object block and the difference of its predicted value mark, And represent whether above-mentioned process object block includes the mark of the coefficient of brightness and aberration.
In addition, the method for encoding images of a technical scheme for the present invention, it is characterised in that in the control of above-mentioned context In step, the also position based on above-mentioned process object block, judge in coding whether can using block on above-mentioned control parameter； The above-mentioned context of above-mentioned 2nd conditional decision can not used in the case of control parameter using block on above-mentioned.
In addition, the method for encoding images of a technical scheme for the present invention, it is characterised in that in the control of above-mentioned context In step, in the case where above-mentioned process object block belongs to slice boundaries, it is judged as that block on above-mentioned can not be utilized in coding Control parameter.
In addition, the method for encoding images of a technical scheme for the present invention, it is characterised in that in the control of above-mentioned context In step, the depth of the level of the data unit according to belonging to the control parameter of above-mentioned process object block judge in coding whether Above-mentioned upper piece control parameter can be utilized.
In addition, the method for encoding images of a technical scheme for the present invention, it is characterised in that in the control of above-mentioned context In step, the value of the control parameter also based on the 1st unit contrasts the control parameter of the 2nd small unit of above-mentioned 1st unit, switching It is to use the above-mentioned context of above-mentioned 1st conditional decision or use above-mentioned 2nd conditional decision context.
In addition, the picture coding device of a technical scheme for the present invention, uses arithmetic coding, it is characterised in that
Including：
Binaryzation portion, passes through the control parameter binaryzation by block is dealt with objects, generation two-value arrangement；
Context control unit, determines the context used the arithmetic coding for dealing with objects block in multiple contexts；With And
Arithmetic coding portion, arithmetic coding, generation are carried out by using the above-mentioned context determined to the arrangement of above-mentioned two-value Bit sequence；
Above-mentioned context control unit,
Judge the signal kinds of the control parameter of above-mentioned process object block；
In the case where above-mentioned signal kinds are the 1st species, using using with above-mentioned process object block abut left piece and 1st condition of upper piece encoded control parameter, determines above-mentioned context；
In the case where above-mentioned signal kinds are with the above-mentioned 1st diverse 2nd species, using it is unfavorable with it is above-mentioned upper piece And above-mentioned left piece encoded control parameter and utilize the 2nd condition of fixed value set in advance, determine above-mentioned context；
Above-mentioned 1st species is to represent whether above-mentioned process object block is divided into multiple pieces of mark or indicate whether will be upper State the mark that process object block is skipped；
Above-mentioned 2nd species be represent the motion vector composition of above-mentioned process object block and the difference of its predicted value mark, And represent whether above-mentioned process object block includes the mark of the coefficient of brightness and aberration.
In addition, the image encoding/decoding device of a technical scheme for the present invention, it is characterised in that including：
Use the picture coding device of arithmetic coding；And
Use the picture decoding apparatus of arithmetic decoding；
Above-mentioned picture coding device includes：
Binaryzation portion, by the way that by the 1st control parameter binaryzation of the 1st process object block, the 1st two-value of generation is arranged；
1st context control unit, determines what the arithmetic coding of the 1st process object block was used in multiple 1st contexts 1st context；And
Arithmetic coding portion, arithmetic volume is carried out by using above-mentioned 1st context determined to the arrangement of above-mentioned 1st two-value Code, generates the 1st bit sequence；
Above-mentioned 1st context control unit,
Judge the 1st signal kinds of the 1st control parameter of above-mentioned 1st process object block；
In the case where above-mentioned 1st signal kinds are the 1st species, deal with objects what block was abutted using using with the above-mentioned 1st 1st condition of the 1st encoded control parameter of block on 1st left piece and the 1st, determines above-mentioned 1st context；
In the case where above-mentioned 1st signal kinds are with the above-mentioned 1st diverse 2nd species, using unfavorable with above-mentioned 2nd condition of block and above-mentioned 1st left piece of the 1st encoded control parameter on the 1st and utilization the 1st fixed value set in advance, Determine above-mentioned 1st context；
Above-mentioned 1st species is to represent whether above-mentioned 1st process object block is divided into multiple pieces of mark or indicates whether The mark that above-mentioned 1st process object block is skipped；
Above-mentioned 2nd species is the motion vector composition and the mark of the difference of its predicted value for representing above-mentioned 1st process object block Whether will and the above-mentioned 1st process object block of expression include the mark of the coefficient of brightness and aberration；
Above-mentioned picture decoding apparatus includes：
2nd context control unit, determines what the arithmetic decoding of the 2nd process object block was used in multiple 2nd contexts 2nd context；
Arithmetic decoding portion, will be by dealing with objects block obtained from the 2nd control parameter progress arithmetic coding with the above-mentioned 2nd Corresponding 2nd bit sequence, uses above-mentioned 2nd context determined to carry out arithmetic decoding, and the 2nd two-value thus is arranged into multiple It is former；And
Intermediate portion, by the way that above-mentioned 2nd two-value is arranged into intermediate, above-mentioned 2nd control parameter is restored；
Above-mentioned 2nd context control unit,
Judge the 2nd signal kinds of the 2nd control parameter of above-mentioned 2nd process object block；
In the case where above-mentioned 2nd signal kinds are the 3rd species, deal with objects what block was abutted using using with the above-mentioned 2nd 3rd condition of decoded 2nd control parameter of block on 2nd left piece and the 2nd, determines above-mentioned 2nd context；
In the case where above-mentioned 2nd signal kinds are with the above-mentioned 3rd diverse 4th species, using unfavorable with above-mentioned 4th condition of block and above-mentioned 2nd left piece of decoded 2nd control parameter on the 2nd and utilization the 2nd fixed value set in advance, Determine above-mentioned 2nd context；
Above-mentioned 3rd species is to represent whether above-mentioned 2nd process object block is divided into multiple pieces of mark or indicates whether The mark that above-mentioned 2nd process object block is skipped；
Above-mentioned 4th species is the motion vector composition and the mark of the difference of its predicted value for representing above-mentioned 2nd process object block Whether will and the above-mentioned 2nd process object block of expression include the mark of the coefficient of brightness and aberration.
Invention effect
The present invention, which can be provided, can cut down the method for encoding images or picture decoding method of memory usage amount.
Brief description of the drawings
Fig. 1 is the functional block diagram of the picture coding device of embodiment 1 for the present invention.
Fig. 2 is the functional block diagram of the variable-length encoding part of embodiment 1 for the present invention.
Fig. 3 is the table of the context model of the control parameter on embodiment 1 for the present invention.
Fig. 4 is the flow chart for the arithmetic coding method for representing embodiment 1 for the present invention.
Fig. 5 is the functional block diagram of the picture decoding apparatus of embodiment 2 for the present invention.
Fig. 6 is the functional block diagram in the variable-length decoding portion of embodiment 2 for the present invention.
Fig. 7 is the flow chart for the arithmetic decoding method for representing embodiment 2 for the present invention.
Fig. 8 is the flow chart of the variation for the arithmetic decoding method for representing embodiment 2 for the present invention.
Fig. 9 A are for illustrating the corresponding figure between the map information and context index of embodiment of the present invention 2.
Fig. 9 B are the figures of the segmentation block (tree construction) for the HEVC for representing embodiment 2 for the present invention.
Figure 10 is for illustrating the figure that the multi-layer block of embodiment 2 for the present invention is constructed.
Figure 11 is the arithmetic decoding method for the split_coding_unit_flag for representing embodiment 3 for the present invention Table.
Figure 12 A are the checking knots on split_coding_unit_flag for representing embodiment 3 for the present invention The table of fruit.
Figure 12 B are the checking knots on split_coding_unit_flag for representing embodiment 3 for the present invention The table of fruit.
Figure 13 is the table of the arithmetic decoding method for the skip_flag for representing embodiment 3 for the present invention.
Figure 14 A are the tables for the result on skip_flag for representing embodiment 3 for the present invention.
Figure 14 B are the tables for the result on skip_flag for representing embodiment 3 for the present invention.
Figure 15 is the table of the arithmetic decoding method for the merge_flag for representing embodiment 3 for the present invention.
Figure 16 A are the tables for the result on merge_flag for representing embodiment 3 for the present invention.
Figure 16 B are the tables for the result on merge_flag for representing embodiment 3 for the present invention.
Figure 17 is the table of the arithmetic decoding method for the ref_idx for representing embodiment 3 for the present invention.
Figure 18 A are the tables for the result on ref_idx for representing embodiment 3 for the present invention.
Figure 18 B are the tables for the result on ref_idx for representing embodiment 3 for the present invention.
Figure 18 C are the tables for the context model on ref_idx for representing embodiment 3 for the present invention.
Figure 18 D are the tables for the context model on ref_idx for representing embodiment 3 for the present invention.
Figure 19 is the table of the arithmetic decoding method for the inter_pred_flag for representing embodiment 3 for the present invention.
Figure 20 A are the tables for the result on inter_pred_flag for representing embodiment 3 for the present invention.
Figure 20 B are the tables for the result on inter_pred_flag for representing embodiment 3 for the present invention.
Figure 21 is the table of the arithmetic decoding method for the mvd for representing embodiment 3 for the present invention.
Figure 22 A are the tables for the result on mvd for representing embodiment 3 for the present invention.
Figure 22 B are the tables for the result on mvd for representing embodiment 3 for the present invention.
Figure 22 C are the tables for the context model on mvd for representing embodiment 3 for the present invention.
Figure 22 D are the tables for the context model on mvd for representing embodiment 3 for the present invention.
Figure 23 A are the arithmetic decoding methods for the no_residual_data_flag for representing embodiment 3 for the present invention Table.
Figure 23 B are the tables of the grammer for the no_residual_data_flag for representing embodiment 3 for the present invention.
Figure 24 A are the results on no_residual_data_flag for representing embodiment 3 for the present invention Table.
Figure 24 B are the results on no_residual_data_flag for representing embodiment 3 for the present invention Table.
Figure 25 A are the arithmetic decoding sides for the intra_chroma_pred_mode for representing embodiment 3 for the present invention The table of method.
Figure 25 B be represent embodiment 3 for the present invention, it is corresponding with intra_chroma_pred_mode The table of IntraPredModeC determining method.
Figure 26 A are the checking knots on intra_chroma_pred_mode for representing embodiment 3 for the present invention The table of fruit.
Figure 26 B are the checking knots on intra_chroma_pred_mode for representing embodiment 3 for the present invention The table of fruit.
Figure 27 is the table of the arithmetic decoding method for the cbf_luma (cr, cb) for representing embodiment 3 for the present invention.
Figure 28 A are the tables for the result on cbf_luma (cr, cb) for representing embodiment 3 for the present invention.
Figure 28 B are the tables for the result on cbf_luma (cr, cb) for representing embodiment 3 for the present invention.
Figure 29 A are the block diagrams for the result for representing embodiment 3 for the present invention.
Figure 29 B are the block diagrams for the result for representing embodiment 3 for the present invention.
Figure 30 A are the block diagrams for the result for representing embodiment 3 for the present invention.
Figure 30 B are the block diagrams for the result for representing embodiment 3 for the present invention.
Figure 31 A are the block diagrams for the result for representing embodiment 3 for the present invention.
Figure 31 B are the block diagrams for the result for representing embodiment 3 for the present invention.
Figure 32 A are the block diagrams for the result for representing embodiment 3 for the present invention.
Figure 32 B are the block diagrams for the result for representing embodiment 3 for the present invention.
Figure 33 A are the tables of one of the parameter set for representing embodiment 3 for the present invention.
Figure 33 B are the tables for the result for representing embodiment 3 for the present invention.
Figure 34 A are the tables of one of the parameter set for representing embodiment 3 for the present invention.
Figure 34 B are the tables for the result for representing embodiment 3 for the present invention.
Figure 35 is for illustrating embodiment for the present invention, the corresponding control parameter using adjacent two blocks The figure of the context model of value.
Figure 36 be for illustrate embodiment for the present invention, using upper piece when memory usage amount it is increased Figure.
Figure 37 is the overall structure figure for the content provider system for realizing content distribution service.
Figure 38 is the overall structure figure of digital broadcasting system.
Figure 39 is the module map for the configuration example for representing television set.
Figure 40 is the configuration example for representing to enter the recording medium as CD information regeneration/record portion of the read-write of row information Module map.
Figure 41 is the figure of the structure example for the recording medium for being denoted as CD.
Figure 42 A are the figures of one for representing portable phone.
Figure 42 B are the module maps for the configuration example for representing portable phone.
Figure 43 is the figure for the structure for representing multiplex data.
Figure 44 is to show schematically each figure for flowing and how being re-used in multiplex data.
Figure 45 is to represent how video flowing is saved in PES packet sequences figure in more detail.
Figure 46 is the figure of the construction of the TS bags and source bag that represent multiplex data.
Figure 47 is the figure for the data structure for representing PMT.
Figure 48 is the figure for the internal structure for representing multiplex data information.
Figure 49 is the figure for the internal structure for representing stream attribute information.
Figure 50 is figure the step of representing to recognize image data.
Figure 51 is the integrated electricity for representing to realize the dynamic image encoding method of each embodiment and dynamic image decoding method The module map of the configuration example on road.
Figure 52 is the figure for the structure for representing switching driving frequency.
Figure 53 is figure the step of representing to recognize image data, switching driving frequency.
Figure 54 is the figure of one for representing to establish the standard of image data and driving frequency corresponding look-up table.
Figure 55 A are represented the figure of one of the shared structure of the module of signal processing part.
Figure 55 B are represented the figure of another of the shared structure of the module of signal processing part.
Embodiment
(being used as the basic understanding of the present invention)
The inventors have found that, it may occur that the problem of following.
Here, in the High-Efficiency Video Coding (HEVC) as coded system of future generation, on The research of context model in the encoding and decoding of various control parameters is also being in progress (non-patent literature 2).So-called control Parameter, refers to be included in coded bit stream, the parameter that uses in coded treatment or decoding process (mark etc.), it is specific and Speech, is syntax element.
So-called context model, is to represent (1) (multivalue, binary numeral or binary number are arranged to what kind of unit Each key element of (bin string)) signal, (2) information of condition (Condition) for giving consideration to how.Here, it is so-called why The condition of sample, is what kind of condition to want the condition of prime number or the signal kinds of the control parameter considered as condition using Preferably what.The condition is distinguished more in detail, i.e. increase conditional number τ (the number of conditions), Number of times corresponding with 1 condition is more reduced.As a result, reduced by the study number of times of each condition, under the precision of the prediction of probability Drop is (for example, referring to " the dilution effect " of non-patent literature 1).
In addition, reducing conditional number, do not consider context (condition of surrounding), do not follow (adaptive) Statistics changes.
, it is necessary to statistics by carrying out picture material after the design guideline of decision model in the modelling of context The checking of fluctuation controls the special for image of the checking of statistical fluctuation of control parameter of encoding and decoding of image etc. The checking of change considers its properness.
In h .264, the event in advance of the quantity limited will be used in the coding of code element as the specification of rule, and will Context model type turns to 4 basic types (basic design types).
Encoding and decoding of 1st and the 2nd type on control parameter.
The context model of 1st type is used with the encoded value (reference of two adjoinings for the upper limit (up to two) Non-patent literature 1).The definition of two adjacent encoded values depends on each signal kinds of control parameter, but generally makes With the value of the corresponding control parameter included in the block of left and upper adjoining.
The type of 2nd context model is the type for determining context based on binary number tree as probability of happening.Tool For body, it is applied in control parameter mb_type and sub_mb_type.
Encoding and decoding of 3rd and the 4th type on the residual values (residual data) of view data etc..In the 3rd class In type, according to the order of the scanning of coefficient of frequency (or quantization parameter), the value for encoding or decoding merely with the past.In the 4th type In, context is determined according to the value (level (level) value) for decoding and accumulating.
Above-mentioned 1st type etc., probability H.264 shift the design guideline of (probability transition) model And installation method demonstrates effect by longer, research is in progress, to be also applied to (ginseng in the HEVC in research at present According to non-patent literature 2).For example studying the 1st type (using adjacent syntactic element (using neighbouring Syntax elements) context model) be used in control parameter alf_cu_flag, split_coding_unit_flag, skip_flag、merge_flag、intra_chroma_pred_mode、inter_pred_flag、ref_idx_lc、ref_ idx_l0、ref_idx_l1、mvd_l0、mvd_l1、mvd_lc、no_residual_data_flag、cbf_luma、cbf_cb And in cbf_cr (9.3.3.1.1 with reference to non-patent literature 2 is saved).
But, on the coding using the 1st type " utilizing adjacent two pieces context model ", the inventors have found that, The problem of having following on its memory usage amount.
Figure 35 is the figure for illustrating the context model of the value using adjacent two pieces corresponding control parameter.In addition, Figure 35 represent H.264 in utilization adjacent block context model.
Block C in figure includes the value as the coding of current (Current) or the control parameter SE of decoder object.Should In the case of SE value coding, the control parameter SE of the same race included in encoded upper piece A and left piece of B value is utilized.Tool For body, use above block A control parameter SE value and left piece of B control parameter SE value for condition conditional probability p (x | (the condition A value of block (upper) and condition B (left piece of value))) prediction block C control parameter SE (or control parameters SE bin string the 1st binary number) value x be " 1 " or " 0 " Probability p (x).
Figure 36 is for illustrating the increased figure using the memory usage amount in the case of upper piece.
(xP, yP) in figure represents the Prediction Unit (PU for including block C：Motion prediction unit) top left pixel Position.Here, block C is the block for the control parameter (such as skip_flag) for including current coded object.In addition, in figure (xP, yA) is denoted as pixel that condition A (the control parameter skip_flag of upper block value) are used, being included in block B Position.In figure (xL, yP) be denoted as it is that condition B (left piece of control parameter skip_flag value) are used, in block A Comprising location of pixels.
In the case, in order to encode or decode block C control parameter skip_flag value, coding or decoding apparatus Need to keep the corresponding PU in position of (xP, the yA) with being included in upper piece B skip_flag value (or the judgement knot of condition Really) and with the corresponding PU in position of (xL, the yP) skip_flag included in left piece of A value (or judged result of condition). Here, if the width for setting the picture is 4096 pixels, in order to 1 control parameter skip_flag coding, it is necessary to keep on The whole judgment values included in block row (the Line L shown in Figure 36).That is, for 1 control parameter, it is necessary to 4096 pixel ÷ The memory span of block size.
Here, the block C of coded object block size is variable, e.g. 64 × 64,16 × 16 or 4 × 4 etc..In addition, When the coding of the block included in the row (Line L) of the top comprising (xP, yA) or decoding, the block C for encoding or decoding after Block size can not predict.Because, the coding or decoding of coding or decoding apparatus row (row for including block A) up Time point, it is not known that each piece of size of the row (row for including block C) under it.Thus, coding or decoding apparatus are contemplated under conduct Side row block size and use using the control parameter block size in minimum block size situation, it is necessary to according to each The minimum block size keeps the value (or judgment value) of control parameter.In addition, the positional representation of Figure 36 black circle is actual in lower section Row (row for including block C) coding or decoding in although not being needed the condition value the condition that must keep.
And then, two pieces of adjoining shown in Figure 36 (left piece of A and upper piece B) represents the concept of the H.264 adjacent block of time point, does not have Import the viewpoint of new level block segmentation.As described later, to being adapted to the piecemeal cut tree structure of recurrence 4 that plan is imported with HEVC The control parameter of (quad tree partitioning) is made, according to the position of recursive execution sequence, hierarchical depth or block, is had By the nonsensical situation of the condition value of Figure 36 references represented.
So, the inventors have found that, by the arithmetic coding of control parameter or decoding utilize upper piece condition value, deposit The increase of reservoir usage amount.And then, the inventors have found that, in HEVC, the memory usage amount further increases.
In this regard, the picture decoding method of a form for the present invention, uses arithmetic decoding, it is characterised in that including：On Hereafter rate-determining steps, determine the context used process object block in multiple contexts；Arithmetic decoding step, will be by right Control parameter carry out arithmetic coding obtained from, bit sequence corresponding with above-mentioned process object block use determined it is above-mentioned on Arithmetic decoding is hereafter carried out, thus two-value is arranged and restored；And intermediate step, by the way that above-mentioned two-value is arranged into intermediate, Above-mentioned control parameter is restored；In above-mentioned context rate-determining steps, the signal of the control parameter of above-mentioned process object block is judged Species；In the case where above-mentioned signal kinds are the 1st species, using using with above-mentioned process object block abut left piece and upper piece Decoded control parameter the 1st condition, determine above-mentioned context；It is different from above-mentioned 1st species in above-mentioned signal kinds The 2nd species in the case of, using unfavorable the 2nd condition with above-mentioned upper piece decoded control parameter, determine above and below above-mentioned Text；Above-mentioned 1st species is " split_coding_unit_flag " or " skip_flag "；Above-mentioned 2nd species is " merge_ Flag ", " ref_idx_l0 " or " ref_idx_l1 ", " inter_pred_flag ", " mvd_l0 " or " mvd_l1 ", " no_ Residual_data_flag ", " intra_chroma_pred_mode ", " cbf_luma ", " cbf_cb " or " cbf_cr ".
Accordingly, the picture decoding method can cut down memory usage amount.Specifically, the picture decoding method is due to right The control parameter of 2nd species is without using upper piece control parameter, so the control parameter of the 2nd species of block need not be kept. Thus, the picture decoding method with it is standardizedly and sharp as " context model of the value of control parameter of the utilization based on adjacent block " Memory usage amount can be cut down by being compared with left piece with upper piece situation.And then, the picture decoding method can not damage image BD-rate values evaluate etc. and suitably cut down the memory usage amount of the control parameter of the 2nd species.
In addition, the picture decoding method can carry out prior art H.264 in do not account for, be suitable for new The utilization or memory reference of the context of the hierarchical tree construction of the distinctive data configurations of HEVC.
In addition it is also possible to be, above-mentioned 2nd condition is unfavorable to be used with above-mentioned upper piece decoded control parameter is also unfavorable State the condition of left piece of decoded control parameter.
Accordingly, the picture decoding method by addition to upper piece also without using left piece of control parameter, can be further Cut down memory usage amount.
In addition it is also possible to which it is the feelings of above-mentioned 2nd species in above-mentioned signal kinds in above-mentioned context rate-determining steps to be Under condition, as above-mentioned 2nd condition, by context set in advance be determined as using in the arithmetic decoding of above-mentioned object block it is upper Hereafter.
Accordingly, the picture decoding method can reduce treating capacity.
In addition it is also possible to be, in above-mentioned context rate-determining steps, the also position based on above-mentioned process object block judges Whether the control parameter of block on above-mentioned can be utilized in decoding；It can not utilize on above-mentioned in the case of the control parameter of block, Use the above-mentioned context of above-mentioned 2nd conditional decision.
Accordingly, the picture decoding method can reduce treating capacity.
In addition it is also possible to be in above-mentioned context rate-determining steps, to belong to the feelings of slice boundaries in above-mentioned process object block Under condition, it is judged as that the control parameter of block on above-mentioned can not be utilized in decoding.
In addition it is also possible to be, in above-mentioned context rate-determining steps, according to belonging to the control parameter of above-mentioned process object block Data unit level depth judge in decoding whether can using block on above-mentioned control parameter.
In addition it is also possible to be, above-mentioned 2nd species is the control parameter with data configuration set in advance.
In addition it is also possible to be, in above-mentioned context rate-determining steps, the value of the control parameter also based on the 1st unit, contrast The control parameter of the 2nd small unit of above-mentioned 1st unit, switching is using the above-mentioned context of above-mentioned 1st conditional decision or used Above-mentioned 2nd conditional decision context.
In addition it is also possible to be, above-mentioned " split_coding_unit_flag " represents whether above-mentioned process object block is divided It is segmented into multiple pieces；Above-mentioned " skip_flag " indicates whether to skip above-mentioned process object block；Above-mentioned " merge_flag " expression pair Whether above-mentioned process object block uses merging patterns；Above-mentioned " ref_idx_l0 " represents the list 0 of above-mentioned process object block Reference picture is indexed；Above-mentioned " ref_idx_l1 " represents the reference picture index of the list 1 of above-mentioned process object block；It is above-mentioned " Which of uni-prediction and bi-prediction inter_pred_flag " expressions use above-mentioned process object block； Above-mentioned " mvd_l0 " represents the motion vector composition and the difference of its predicted value of the list 0 used above-mentioned process object block；On State the motion vector composition and the difference of its predicted value for the list 1 that " mvd_l1 " expression is used above-mentioned process object block；It is above-mentioned " No_residual_data_flag " indicates whether the residual error data that there is above-mentioned process object block；Above-mentioned " intra_ Chroma_pred_mode " represents the intra prediction mode of the aberration sample of above-mentioned process object block；Above-mentioned " cbf_luma " Whether represent the luminance transformation block of above-mentioned process object block includes the more than one conversion coefficient that rank is not 0；Above-mentioned " cbf_ Cb " represents whether the Cb transform blocks of above-mentioned process object block include the more than one conversion coefficient that rank is not 0；It is above-mentioned " Cbf_cr " represents whether the Cr transform blocks of above-mentioned process object block include the more than one conversion coefficient that rank is not 0.
In addition it is also possible to be, above-mentioned picture decoding method is according to included in encoded signal, the 1st standard of expression or the 2nd mark Accurate identifier, is switched over to the decoding process according to above-mentioned 1st standard and according to the decoding process of above-mentioned 2nd standard； In the case that above-mentioned ID represents the 1st standard, as the decoding process according to above-mentioned 1st standard, above-mentioned context control step is carried out Suddenly, above-mentioned arithmetic decoding step and above-mentioned intermediate step.
In addition, the method for encoding images of a form for the present invention, uses arithmetic coding, it is characterised in that including：Two Value step, passes through the control parameter binaryzation by block is dealt with objects, generation two-value arrangement；Context rate-determining steps, multiple In context, the context used above-mentioned process object block is determined；And arithmetic coding step, it is upper by using what is determined State context and arithmetic coding is carried out to the arrangement of above-mentioned two-value, generate bit sequence；In above-mentioned context rate-determining steps, in judgement State the signal kinds of the control parameter of process object block；In the case where above-mentioned signal kinds are the 1st species, using using with it is upper Left piece of process object block adjoining and the 1st condition of upper piece control parameter are stated, above-mentioned context is determined；In above-mentioned signal kinds Be with the above-mentioned 1st diverse 2nd species in the case of, using unfavorable the 2nd condition with above-mentioned upper piece control parameter, certainly Fixed above-mentioned context；Above-mentioned 1st species is " split_coding_unit_flag " or " skip_flag "；Above-mentioned 2nd species It is " merge_flag ", " ref_idx_l0 " or " ref_idx_l1 ", " inter_pred_flag ", " mvd_l0 " or " mvd_ L1 ", " no_residual_data_flag ", " intra_chroma_pred_mode ", " cbf_luma ", " cbf_cb " or " cbf_cr"。
Accordingly, the method for encoding images can cut down memory usage amount.Specifically, the method for encoding images is due to right The control parameter of 2nd species is without using upper piece control parameter, so the control parameter of the 2nd species of block need not be kept. Thus, the method for encoding images with it is standardizedly and sharp as " context model of the value of control parameter of the utilization based on adjacent block " Memory usage amount can be cut down by being compared with left piece with upper piece situation.And then, the method for encoding images can not damage image BD-rate values evaluate etc. and suitably cut down the memory usage amount of the control parameter of the 2nd species.
In addition, the method for encoding images can carry out prior art H.264 in do not account for, be suitable for new The utilization or memory reference of the context of the hierarchical tree construction of the distinctive data configurations of HEVC.
In addition, the picture decoding apparatus of a form for the present invention, uses arithmetic decoding, it is characterised in that including：On Hereafter control unit, determines the context used process object block in multiple contexts；Arithmetic decoding portion, will be by control Parameter carry out arithmetic coding obtained from, bit sequence corresponding with above-mentioned process object block use the above-mentioned context determined Arithmetic decoding is carried out, thus two-value is arranged and restored；And intermediate portion, will be above-mentioned by the way that above-mentioned two-value is arranged into intermediate Control parameter is restored；Above-mentioned context control unit, judges the signal kinds of the control parameter of above-mentioned process object block；In above-mentioned letter In the case that number species is the 1st species, use using and above-mentioned process object block left piece and upper piece decoded control abutting 1st condition of parameter processed, determines above-mentioned context；It is the feelings with the above-mentioned 1st diverse 2nd species in above-mentioned signal kinds Under condition, using unfavorable the 2nd condition with above-mentioned upper piece decoded control parameter, above-mentioned context is determined；Above-mentioned 1st species It is " split_coding_unit_flag " or " skip_flag "；Above-mentioned 2nd species is " merge_flag ", " ref_idx_ L0 " or " ref_idx_l1 ", " inter_pred_flag ", " mvd_l0 " or " mvd_l1 ", " no_residual_data_ Flag ", " intra_chroma_pred_mode ", " cbf_luma ", " cbf_cb " or " cbf_cr ".
Accordingly, the picture decoding apparatus can cut down memory usage amount.
In addition, the picture coding device of a form for the present invention, uses arithmetic coding, it is characterised in that including：Two Value portion, passes through the control parameter binaryzation by block is dealt with objects, generation two-value arrangement；Context control unit, above and below multiple Wen Zhong, determines the context used above-mentioned process object block；And arithmetic coding portion, by using determined it is above-mentioned above and below Text carries out arithmetic coding to the arrangement of above-mentioned two-value, generates bit sequence；Above-mentioned context control unit, judges above-mentioned process object block Control parameter signal kinds；In the case where above-mentioned signal kinds are the 1st species, utilization and above-mentioned process object block are used Left piece adjacent and the 1st condition of upper piece control parameter, determines above-mentioned context；It is and the above-mentioned 1st in above-mentioned signal kinds In the case of diverse 2nd species, using unfavorable the 2nd condition with above-mentioned upper piece control parameter, determine above and below above-mentioned Text；Above-mentioned 1st species is " split_coding_unit_flag " or " skip_flag "；Above-mentioned 2nd species is " merge_ Flag ", " ref_idx_l0 " or " ref_idx_l1 ", " inter_pred_flag ", " mvd_l0 " or " mvd_l1 ", " no_ Residual_data_flag ", " intra_chroma_pred_mode ", " cbf_luma ", " cbf_cb " or " cbf_cr ".
Accordingly, the picture coding device can cut down memory usage amount.
In addition, the image encoding/decoding device of a form for the present invention includes above-mentioned picture decoding apparatus and above-mentioned figure As code device.
In addition, these master or specific forms can also by system, method, integrated circuit, computer program or Recording medium realization, can also be real by the arbitrary combination of system, method, integrated circuit, computer program and recording medium It is existing.
Hereinafter, referring to the drawings the picture decoding apparatus and picture coding device to a form for the present invention specifically It is bright.
In addition, embodiments described below all represents a specific example of the present invention.In the following embodiments The numerical value of expression, shape, material, inscape, the allocation position of inscape and connection form, step, order of step etc. It is one, is not the meaning for limiting the present invention.In addition, in the inscape of following embodiment, most upper in expression There is no the inscape recorded in the independent claims of position concept, be set to arbitrary inscape and illustrate.
(embodiment 1)
Hereinafter, the picture coding device of embodiment 1 for the present invention is illustrated.Embodiment party for the present invention The picture coding device of formula 1, according to the signal kinds of control parameter, switches above and below (1) utilization upper piece decision in arithmetic coding Text, still (2) are unfavorable with a piece decision context.Used thereby, it is possible to cut down memory while the deterioration of image quality is suppressed Amount.
First, the structure of the picture coding device of embodiment 1 for the present invention is illustrated.
Fig. 1 is the block diagram of the picture coding device 100 about present embodiment.
Picture coding device 100 shown in Fig. 1 is the picture coding device using arithmetic coding, by the way that input picture is believed Numbers 121 codings and generate bit stream 124.The picture coding device 100 includes control unit 101, difference portion 102, conversion and quantified It is pre- between portion 103, variable-length encoding part 104, re-quantization and inverse transformation portion 105, adder 106, intra-frame prediction portion 107, picture Survey portion 108 and switch 109.
Received image signal 121 of the control unit 101 based on coded object calculates control parameter 130.For example, control parameter 130 include information, the motion prediction unit of coded object block of the picture/mb-type of the received image signal 121 of presentation code object The size of (Prediction Unit PU) and the control information of motion prediction unit etc..Here, (the Control of control parameter 130 Data) its own is coded object.Thus, control unit 101 exports the control parameter 130 to variable-length encoding part 104.
Difference portion 102 calculates the difference (residual error of the received image signal 121 and prediction image signal 129 as block unit Value) residual signals 122.
Residual signals 122 are transformed to frequency coefficient values by conversion and quantization unit 103, pass through the frequency coefficient values amount that will be obtained Change, generating quantification conversion coefficient 123 (residual data).
The re-quantization of quantization transform coefficient 123 is frequency coefficient values by re-quantization and inverse transformation portion 105, by by obtained frequency Rate coefficient value inverse transformation, generates the residual signals 125 restored.
Adder 106 is by the way that residual signals 125 are added with prediction image signal 129, output restored image signal 126.
Intra-frame prediction portion 107 carries out intra-frame prediction processing by using restored image signal 126 and generates prognostic chart As signal 127.Inter-picture prediction portion 108 carries out inter-picture prediction processing by using restored image signal 126 and generates prediction Picture signal 128.
Switch 109 selects a sides of prediction image signals 127 and prediction image signal 128, using selected signal as Prediction image signal 129 is exported.
Variable-length encoding part 104 is above-mentioned by the way that each piece of the quantization transform coefficient 123 and control parameter 130 of input are used CABAC coding and generate bit stream 124.
Then, the structure of variable-length encoding part 104 is illustrated.
Fig. 2 is the functional block diagram of variable-length encoding part 104.Variable-length encoding part 104 includes binaryzation portion 141, context Control unit 142 and two-value arithmetic coding portion 143.Hereinafter, the Variable Length Code processing to control parameter 130 is illustrated.In addition, Variable Length Code processing on quantization transform coefficient 123 is omitted the description, for example, known technology can be used to realize.
Binaryzation portion 141 is by the way that by the binaryzation of control parameter 130, generation two-value arranges 151.Specifically, binaryzation portion 141 be perform non-patent literature 1 in II.1) binarizaion processing processing unit.The binaryzation portion 141 is by control parameter 130 two-values for being transformed to referred to as bin string by the binary processing method set according to each signal kinds in advance are arranged 151.In addition, being described below on signal kinds with the corresponding of binary processing method.In addition, binaryzation portion 141 is in input Control parameter 130 be flag etc. 1 binary numeral in the case of, the former state of control parameter 130 is arranged as two-value 151 outputs.
Context control unit 142 is determined in the block of process object is contained in multiple contexts (Probability State table) Control parameter 130 arithmetic coding in the context that uses.In addition, context control unit 142 will specify be determined above and below The context index 152 of text is exported to two-value arithmetic coding portion 143.
Specifically, context control unit 142 is that the 2) contextmodeling performed in non-patent literature 1 is handled Processing unit.For the context control unit 142, sequentially input and wrapped in the two-value arrangement 151 that two-value arithmetic coding portion 143 is exported The multiple key elements contained.Context control unit 142 is arranged according to the signal kinds of control parameter 130 and the two-value of the binary number Elements position in 151, selects the context used in the binary number in multiple contexts, will represent on selected Index hereafter is that context index 152 is exported to two-value arithmetic coding portion 143.
In addition, context control unit 142 is remain as the state of context by the two-value arrangement of control parameter 130 Each key element further distinguishes the Probability State of the quantity (so-called context index number) for the condition for turning to conditional probability in detail Table, the Probability State table is initialized and updated.
In addition, context control unit 142 (wants prime number to be more than 2 as according to each signal kinds what two-value was arranged In the case of, it is each key element number of the two-value arrangement of control parameter 130.It is same as below) distinguish in further detail and keep each Occurrence condition τ (each context) state (probability state index：Probability State is indexed).The state is " 0 " Or probability of happening P (the interior point of ratio of the relatively low side of probability in the two-value of " 1 "：The typically value of 6 bits), it is higher with probability One side is total 7 bit values of which value (1 bit).In addition, so-called hold mode, is initialization and renewal.For example, so-called Renewal is handled, with H264 again it is transformation between 64 limited states, is which Probability State will be now arranged in Probability state (i.e. which probability) indexing changes.
Specifically, the event X hairs of the higher most probable sides of probability of the context control unit 142 in two-value In the case of life, make the ratio of the probability as most probable sides increases slightly.For example, context control unit 142 passes through Make the value increase and decrease 1 of probability index (probability state index) corresponding with 64 tables, can make as most The ratio of the probability of probable sides increases slightly.On the other hand, in (opposite with the probability of prediction) and the relatively low event of probability In the case that not X occur, context control unit 142 makes the ratio of the most probable of holding probability become big, based on rule Fixed zoom factor α (Li such as ≒ 0.95) reduce (with reference to non-patent literature 1, Fig. 6).The context control unit of present embodiment 142 with it is H.264 same, state is changed and is kept based on corresponding table index change values are established, so as to corresponding in view of The change of the α.
Two-value arrangement 151 is calculated by using the context determined by context control unit 142 in two-value arithmetic coding portion 143 Art is encoded, generation bit stream 124 (bit sequence).
Specifically, two-value arithmetic coding portion 143 is the 3) binaryarithmetic for performing non-patent literature 1 The processing unit of coding processing.The two-value arithmetic coding portion 143 is by using the context specified by context index 152 by two Value 151 arithmetic codings of arrangement, so as to generate bit stream 124.Here, so-called arithmetic coding, refers on various signal kinds The event that occurs of control parameter 130 as the accumulative disposal of probability, what kind of event one side constriction scope will be carried out while with 1 Prescribed limit on number line (number line) sets up correspondence.
First, two-value arithmetic coding portion 143 by 1 number line according to provided from context control unit 142, binary number energy It is divided into two half-intervals in the probability of happening of two values enough taken.The value (such as " 0 ") of the binary number actually occurred be compared with In the case of the value of high probability (probability (such as 0.75) more than 0.5) side, two-value arithmetic coding portion 143 does not change and maintained The lower value limit value low of scope in number line, the result that 1 zoom factor 0.95 is multiplied by with the probability 0.75 to this is corresponding Value be set as new amplitude Range.On the other hand, the binary number actually occurred value be prediction relatively low probability side Value in the case of, two-value arithmetic coding portion 143 makes the lower limit low of scope move the amount of higher probability, makes amplitude Range Changed according to relatively low probability.So, keep interval by probability amplitude Range adding up for multiplied result, but general In the case that the relatively low value of rate is recurred, amplitude Range length turns into below the precision being able to ensure that in computing.Herein In the case of, processing (renorm) and general that amplitude Range expands enter to be about to maintain precision in two-value arithmetic coding portion 143 For the bit sequence output for the scope for representing current time point.On the contrary, in the feelings that recur of value of probability higher (0.95 etc.) Under condition, even if being multiplied by the probable value, amplitude Range length become it is shorter than defined length before can also carry out it is multiple Computing (being state transformation in the case of the installation based on table).Thus, in the case, it can tire out before bit is exported The code element number of meter is more.
Fig. 3 is to arrange the control parameter 130 using the context model of the value of the control parameter 130 based on adjacent block Table.
From the meaning for from left to right, illustrating row.
(c2) signal kinds (syntax element) represent the specific name of the signal kinds of control parameter 130.In addition, The meaning of each signal kinds is aftermentioned.
(c3) binaryzation mode (binarization scheme) represents the control that the row in the left by being close to are specified The binaryzation mode of application in parameter 130 (SE).In addition, binary conversion treatment is performed by above-mentioned binaryzation portion 141.In addition, so-called In column " fixed length ", refer to binaryzation portion 141 by the fixed long (Fixed of the value of the control parameter 130 for the left being close to Length two-value arrangement (bin string) output).Signal kinds name " flag " signal kinds terminated in HEVC Control parameter 130 is 1 binary numeral of some value for taking " 0 " or " 1 ".Thus, the two-value arrangement that binaryzation portion 141 is exported 151 key element is the 1st key element (binIdx=0), the 2nd later key element (binIdx>=1 key element is not exported.That is, Binaryzation portion 141 exports the value of control parameter 130 as two-value arrangement 151 as former state.
In addition, in column " variable length ", binaryzation portion 141 is represented by the value of control parameter 130, using with as with the value The mode of the binary number length of Frequency order from short to long establishes the binary system ordered series of numbers (bin of corresponding variable length String or two-value arrangement, Su Shuo≤1) to carry out two-value arrangementization and export.For example, binaryzation portion 141 will (Truncated) Unary types or Unary types with it is the combination of multiple modes of the Golomb modes of other indexes etc. etc., corresponding Exported (with reference to non-patent literature 1, A.Binarization) in the mode of signal kinds.In addition, in the case of variable length, The arrangement key element for having two-value to arrange 151 is the situation of 1, but also promising more than two situations.In addition, in image solution described later In the intermediate portion of code device, by carrying out the inverse transformation of the binaryzation mode, by the arrangement of the two-value of input be restored to multivalue or Value of statistical indicant.
(c4) the 1st key element (binIdx=0) context index represents 142 pairs of context control unit by by c3 column The choosing of the context index (increment) of the 1st key element application included in the binary number arrangement that the binaryzation mode specified is generated Select branch.Recorded in column " in the case of 0,1,2 ", it is meant that context control unit 142 from 3 Probability State tables (up and down Text) in selection using Probability State table.For example, for signal kinds " skip_flag ", on 1 signal kinds standard Standby 3 context indexs with condition sectionalization, that is, mean that preparing 3 contexts carries out arithmetic coding.
Equally, recorded in c4 columns " in the case of 0,1,2,3 ", it is meant that in the signal kinds that will be specified by c2 columns The value of control parameter 130 is applied to is arranging the included in 151 the 1st by the two-value of the binaryzation mode two-value arrangementization on c3 columns Context in key element (binIdx=0) is that 4 of 0,1,2 or 3 select one.In addition, being described below on the conditional in column.
(c5) left piece of condition L (condL) represents that context control unit 142 is used for selecting certain of 0,1,2 value in column c4 The condition (condition) of left piece of individual value.The condition corresponds to the control parameter of coded object (or decoder object), takes basis True or false value that the value of left piece of control parameter is determined.
For example, in the case where control parameter (SE) is skip_flag, if skip_flag [xL] [yL] value is True (such as " 1 ") then exports true, if false (such as " 0 ") then exports false.
(c6) block condition A (condA) represents that context control unit 142 is used for the volume in the arrangement key element specified by column c4 on Upper piece the condition (condition) of some value of the value of selection 0,1,2 in code and decoding.The condition take with coded object (or Decoder object) control parameter it is corresponding, according to the value of upper piece control parameter determine true or false value.For example, It is defeated if skip_flag [xA] [yA] value is true (such as " 1 ") in the case that control parameter (SE) is skip_flag Go out true, if false (such as " 0 ") then exports false.
In addition, though do not illustrate, but will be to (c7) " binIdx>It is more than the context increment size of=1 application " and 2 bits Signal kinds establish correspondence.Should the binary system later to the 2nd key element of two-value arrangement of (c7) expression context control unit 142 Number (has binIdx>The binary number of=1 index value arranges the binary numeral of key element) context model of application.
About present embodiment method for encoding images for above-mentioned left piece of condition L and upper piece condition A, according to control parameter 130 signal kinds switch following action (utilizing different Modal actions).
(pattern 1) uses two adjacent blocks (left piece of condition L judgment value and upper piece condition A judgment value).
(pattern 2) uses 1 adjacent block (only left piece condition L judgment value).
(pattern 3) uses 0 adjacent block (left piece of condition L and upper piece condition A all without using).
Fig. 4 is the method for encoding images for representing the relevant present embodiment that the variable-length encoding part 104 shown in Fig. 2 is performed Flow chart.
First, binaryzation portion 141 by the value of control parameter 130 with side corresponding with the signal kinds of the control parameter 130 Formula two-value arrangementization (S101).
Then, context control unit 142 obtains the basic of the context that is used in the arithmetic coding of the control parameter 130 It is worth (S102).For example, context control unit 142 determines the basic value according to picture/mb-type (I, P, B).
Then, context control unit 142 uses above-mentioned 1~pattern of pattern 3 according to the signal kinds of control parameter 130 Some determines context value (S103).Here, so-called decision context value, the tune relative to basic value with determining context Whole value (increment value CtxIdxInc) is of equal value.
First, context control unit 142 differentiates the signal kinds (S103) of control parameter 130.In the letter of control parameter 130 Number species is that in the case of the 1st species corresponding with pattern 1 (the 1st species in S104), context control unit 142 is using according to neighbour Judgment value derived from the value of the respective control parameter of two blocks (block A and block B) connect determines context value (S105).Change speech It, context control unit 142 is using using left piece and the conditions of control parameter of upper piece two adjacent blocks determines context. In the case, context control unit 142 utilizes the judged result of (c5) CondL shown in Fig. 3 and (c6) condA judgement knot Both values of fruit.Thus, the control parameter for the 1st species keeps the data that the horizontal stroke one of picture is arranged.
On the other hand, in the case of being the 2nd species corresponding with pattern 2 in the signal kinds of control parameter 130 (in S104 2nd species), context control unit 142 determines the control parameter using 1 adjacent block (the nearest adjacent block on coded sequence) Value determine context value (S106).In other words, context control unit 142 using it is unfavorable with piece the condition of control parameter determine Determine context.
On the other hand, in the case of being the 3rd species corresponding with pattern 3 in the signal kinds of control parameter 130 (in S104 3rd species), 142 pairs of upper blocks of context control unit and left piece of control parameter are not utilized, and regularly determine context value (S107)。
Then, context control unit 142 is added in step by the basic value of the context to determining in step s 102 The increment size determined in S103, export context index value (S108).
Finally, two-value arithmetic coding portion 143 is by the way that the binary numeral of the 1st key element is used by being determined in step S108 The context value specified of context index value, carry out arithmetic coding, generation bit sequence (bit stream 124) (S109).
Then, the whole key elements included in the execution of step S102~S109 processing is arranged two-value are not completed In the case of (no in S110), during variable-length encoding part 104 is arranged two-value the next key element that includes perform step S102~ S109 processing.On the other hand, the whole key elements included in arranging two-value complete step S102~S109 processing Execution in the case of ("Yes" in S110), variable-length encoding part 104 terminate for deal with objects block control parameter coding Processing.
As above, about present embodiment picture coding device 100 in arithmetic coding, the control to the 1st species Parameter processed determines context using upper piece, it is unfavorable to the control parameter of the 2nd species and the 3rd species with piece and determine context.
By the structure, the picture coding device 100 and standardizedly as the " value of control parameter of the utilization based on adjacent block Context model " and memory usage amount can be cut down by being compared using left piece with upper piece situation.Thus, the Image Coding is filled Memory usage amount can be cut down while the deterioration of image quality is suppressed by putting 100.
(embodiment 2)
In the present embodiment, the image decoding to the bit stream 124 generated by above-mentioned picture coding device 100 is decoded Device is illustrated.
Fig. 5 is the block diagram of the picture decoding apparatus 200 about present embodiment.The picture decoding apparatus 200 is using calculation The picture decoding apparatus of art decoding, picture signal 229 is generated by the way that bit stream 124 is decoded.Here, bit stream 124 is for example It is the bit stream 124 generated by above-mentioned picture coding device 100.
Picture decoding apparatus 200 include control unit 201, variable-length decoding portion 202, re-quantization portion 204, inverse transformation portion 205, Adder 206, intra-frame prediction portion 207 and inter-picture prediction portion 208.
Picture decoding apparatus 200 carries out decoding process according to each code sequence of the processing unit of regulation.Here, handle Unit is, for example, section (slice) unit or block unit.
Variable-length decoding portion 202 is by carrying out arithmetic decoding, generation control parameter 230 (control to bit stream 124 Data syntax element) and quantization transform coefficient 223 (Residual data syntax element values).It will give birth to Into control parameter 230 exported to control unit 201.
The processing unit that control unit 201 controls to include in picture decoding apparatus 200 according to control parameter 230.
Re-quantization portion 204 is by by the re-quantization of quantization transform coefficient 223, generating orthogonal transform coefficient 224.
Inverse transformation portion 205 is by the way that by the inverse transformation of orthogonal transform coefficient 224, residual signals 225 are restored.Adder 206 is led to Cross and be added residual signals 225 with prediction image signal (picture signal 229), generate decoded image signal 226.
Intra-frame prediction portion 207 carries out intra-frame prediction processing by using decoded image signal 226 and generates prognostic chart As signal 227.Inter-picture prediction portion 208 carries out inter-picture prediction processing by using decoded image signal 226 and generates prediction Picture signal 228.
Switch 209 selects a sides of prediction image signals 227 and prediction image signal 228, using selected signal as Picture signal 229 (prediction image signal) is exported.
Then, the structure in variable-length decoding portion 202 is illustrated.
Fig. 6 is the functional block diagram for the structure for representing variable-length decoding portion 202.Variable-length decoding portion 202 includes two-value arithmetic Lsb decoder 243, context control unit 242 and intermediate portion 241.Hereinafter, progress is handled to the variable-length decoding of control parameter 230 Explanation.In addition, the variable-length decoding processing on quantization transform coefficient 223 is omitted the description, for example, it can use known technology Realize.
Context control unit 242 determines the arithmetical solution in the control parameter 230 of the block of process object in multiple contexts The context used in code.In addition, context control unit 242 will specify the context index 252 to two of determined context Value arithmetic decoding portion 243 is exported.
Specifically, context control unit 242 is used and the identical context mould of context control unit 142 shown in Fig. 2 Type is used as the probability metastasis model kept.In the case of using 64 Probability States in two-value arithmetic coding portion 143, two-value is calculated Art lsb decoder 243 also has 64 Probability States.Because, it is necessary to by the scope on the number line of coding in coding side and decoding Similarly explain the both sides of side.Thus, also use and selected with coding side from 3 patterns of above-mentioned pattern 1~3 in decoding apparatus side The pattern identical pattern selected.
Two-value arithmetic decoding portion 243 is by using the context determined by context control unit 242 by bit sequence (bit 124) arithmetic decoding is flowed, two-value arrangement 251 is restored.Specifically, two-value arithmetic decoding portion 243 is used by from context control The context (Probability State table) that the context index that portion 242 is provided is specified, two-value arrangement is restored to by the bit sequence of input 251。
Two-value arrangement 251 is carried out intermediate so as to be restored to control parameter 230 by intermediate portion 241 as needed.So, The context control unit 242 that the context control unit 142 and picture decoding apparatus 200 that picture coding device 100 possesses possess exists Identical context model is utilized when the arithmetic coding and arithmetic decoding of the control parameter of some signal kinds in both.
Fig. 7 is the flow chart for the picture decoding method for representing the relevant present embodiment that variable-length decoding portion 202 is performed.
First, variable-length decoding portion 202 obtains bit stream 124 (S201).
Then, context control unit 242 determines the letter of the control parameter of decoder object according to the data configuration of bit stream 124 Number species (S202).
Then, context control unit 242 determines the context used in the arithmetic decoding of the control parameter of decoder object Basic value (S203).For example, context control unit 242 determines the basic value according to picture/mb-type (I, P, B).
Then, context control unit 242 uses some of above-mentioned 1~pattern of pattern 3 according to the signal kinds of control parameter Determine context value (S204).Here, so-called decision context value, the adjusted value relative to basic value with determining context (increment value CtxIdxInc) is of equal value.For example, signal kinds of the context control unit 242 according to control parameter, according to statically thing The table of preceding decision carries out which judgement of 1~pattern of use pattern 3.
Context control unit 242 switches in obtain two-value by arithmetic decoding according to the signal kinds of control parameter The adjacent block used is searched in the decision for the context for arranging the value of the binary number of the 1st key element included in 251 and using.
First, context control unit 242 judges the signal kinds (S205) of control parameter 230.It is and mould in signal kinds In the case of corresponding 1st species of formula 1 (the 1st species in S205), context control unit 242 is respective using two adjacent blocks Control parameter determine context value (S206).In other words, context control unit 242, which is used, utilizes left piece and upper piece two neighbours The condition for connecing the decoded control parameter of block determines context.
On the other hand, in the case where signal kinds are the 2nd species corresponding with pattern 2 (the 2nd species in S205), up and down Literary control unit 242 determines context value using the value of the control parameter of 1 adjacent block (the nearest adjacent block on coded sequence) (S207).In other words, context control unit 242 using it is unfavorable with piece decoded control parameter condition come above and below determining Text.
On the other hand, in the case where signal kinds are the 3rd species corresponding with pattern 3 (the 3rd species in S205), up and down Literary control unit 242 regularly determines context value (S208).In other words, context control unit 242 is using to upper piece and left piece Condition that decoded control parameter is not utilized determines context.
Then, context control unit 242 by the basic value of context index that will be determined in step S203 with step The increment size determined in rapid S204 is added, and determines context index value (S209).
Then, two-value arithmetic decoding portion 243 uses what is represented by the context index value provided from context control unit 242 Context value, 1 (S210) of the key element that two-value is arranged is obtained by decoding.
Then, the whole key elements included in the execution of step S203~S210 processing is arranged for two-value are not complete In the case of ("No" in S211), the next key element that variable-length decoding portion 202 is included in arranging two-value performs step S203 ~S210 processing.
On the other hand, the whole key elements included in arranging two-value complete holding for step S203~S210 processing In the case of row ("Yes" in S211), intermediate portion 241 is by by the place by being repeated 1 times above above-mentioned steps S203~S210 The key element of more than 1 of two-value arrangement 251 is intermediate obtained from reason, generation control parameter 230 (S212).
According to the above, about present embodiment picture decoding apparatus 200 in arithmetic decoding, the control to the 1st species Parameter using upper piece determine context, for the 2nd species and the 3rd species control parameter it is unfavorable with piece and determine context.
By the structure, the picture decoding apparatus 200 and standardizedly as the " value of control parameter of the utilization based on adjacent block Context model " and memory can be cut down and use by being compared using left piece with upper piece situation.Thus, the picture decoding apparatus 200 can cut down memory usage amount while the deterioration of image quality is suppressed.
In addition, intermediate portion 241 is in the case of two-value arrangement 251 is flag etc., wants prime number to be 1, i.e., 1binary's In the case of, two-value arrangement 251 can also as former state be exported.
In addition, in addition to described above, control unit 101 or 201 can also be controlled via signal wire (not shown) The processing of each processing unit or the processing of value with reference to memory etc..
In addition, in the above description, context control unit 142 or 242 is according to the signal kinds switch mode 1 of control parameter 3 patterns of~pattern 3, but it is also possible to two patterns in 1~pattern of signal kinds switch mode 3.In other words, up and down Literary control unit 142 or 242 can also be switched over according to the signal kinds of control parameter to utilization/unfavorable with a piece condition.
In addition, context control unit 142 or 242 can also by the switching method of the context model of such selection ( Situation including changing context model increment size, same as below) changed according to the image information of regulation.For example, context control Portion 142 or 242 processed can also be according to the width of image influential on the study number of times of memory maintenance dose and each context Size or sample format etc. further switch the switching policy in itself.
In addition, in the above description, to illustrate schematically change, context control unit 142 or 242 pairs of utilizations/do not utilize Upper piece condition is switched over, but context control unit 142 or 242 can also be in combination by upper piece unavailable situation of script And apply.For example, whether context control unit 142 or 242 can also be entropy slices (entropy_ according to the section of process object Slice_flag is 1 or 0) switches the switching policy in itself.Equally, originally it cannot be guaranteed that the utilizability of upper adjacent block In the case of, context control unit 142 or 242 can also it is unfavorable with piece and change switching policy.
For example, as shown in figure 8, context control unit 142 or 242 can also be according to the value of the parameter of the unit of regulation (S301) the decision policy of context model, is determined that specification (S302) and the 2nd determines that specification (S303) switches by the 1st.This In, the value of the parameter of the so-called unit according to regulation, as described above, be according to section whether be entropy slice etc..This Outside, the so-called 1st specification is determined, is the specification for carrying out the processing shown in Fig. 7.So-called 2nd determines specification, is not included shown in Fig. 7 Step S204 decision specification, e.g. prior art decision specification.This by the increment size of context index with passing through office It is of equal value that the value of the parameter of the parameter of the defined unit in portion and the unit bigger than the unit, which is determined,.
That is, context control unit 142 or 242 can also the value based on the control parameter of the 1st unit, will be than the 1st unit The decision specification applied in small unit switches to other to determine specification.
Fig. 9 A are the figures for illustrating the correspondence (assignment) between above-mentioned map information and context index.Fig. 9 A The middle example for representing signal mvd_l0, l1, lc.In addition, being also same for other signal kinds.
Assignment 901B shown in Fig. 9 A is the assignment of the context index used in non-patent literature 1.14 deviants 0~ 13 distribute to P pictures.Also, 14 deviants 14~27 are assigned to B pictures.In addition, mvd_l0 [] [] [0] and mvd_l0 [] [] [1] is each signal component value (horizontal direction and vertical direction) in the difference of motion vector respectively.In current research In HEVC, binary context condition value as the 1st key element (binIdx=0) for calculating two-value arrangement (passes through CondA and condL and the condition being refined), be assigned 3 deviants 0~2,3 deviants 7~9,3 deviants 14~ 16 and 3 deviants 21~23.Also, the relation between signal kinds and context index is unrelated with each image sequence And it is fixed to be set to.
Assignment 902B~904B of context index shown in Fig. 9 A is the assignment of the context index value of present embodiment.
The distribution (allocation) of context index when assignment 902B represents use pattern 2 (unfavorable with piece).This In the case of kind, 3 deviants 0~2 etc. need not be distributed as condition value, deviant 0 and 1 the two be enough.Because not Utilize condA.It therefore, there is no need to carry out the distribution of context index for the part of the hacures shown in Fig. 9 A.Therefore, even if To a binIdx in the same manner as assignment 901B>The 0 distribution context index of each 4, total context index be 0~23 this 24 are enough.Therefore, it is possible to cut down at least four context.
The distribution of context index when assignment 903B represents use pattern 3 (unfavorable with piece and left piece of this both sides).This Kind in the case of, need not distribute 0~2 grade 3 as condition value, only distribution 0 this be enough.Because not utilizing CondA does not also utilize condL.It therefore, there is no need to carry out the part of the hacures shown in Fig. 9 A the assignment of context index.Cause This, total context index be 0~19 this 20 be enough.Therefore, it is possible to cut down at least eight context.
Assignment 904B is that, as the big unit of the unit of the block than foregoing signal kinds, image sequence does not include B pictures And the example of the assignment utilized in the situation (situation of front reference is used only) constituted.In this case, as context Index, avoids the need for utilizing the context index for B pictures originally.
Therefore, as context index (relative value), as shown in Figure 9 A, 0~9 these three be enough.Therefore, it is possible to cut down At least 18 contexts.
Also, switching standards illustrated in fig. 8 may also mean that, for all or part of image sequence (predetermined unit), corresponding to the species of its parameter, assignment 901B~904B used in switching.
So, it according to present embodiment, can be indexed according to multiple standards come switch contexts, and be in the prior art root Context index is distributed according to static 1 standard (using upper piece standard with left piece).Therefore, memory can not only be cut down Usage amount, additionally it is possible to according to the property of the unit of regulation, the distribution policy of the context needed for only switching.
In addition, context control unit 142 or 242 can also be changed according to the feature of image sequence used in determine rule Model.For example, context control unit 142 or 242 can also be changed according to the interval (IntraPeriod setting value) of I pictures Used decision specification.
In addition, context control unit 142 or 242 determines specification according to above-mentioned condition switching here, but it is also possible to which switching is It is no to utilize upper piece.
In addition, context control unit 142 or 242 can also be based on above whether the control parameter of block can compiled in position Utilized when code or decoding, to decide whether to utilize upper piece control parameter.That is, context control unit 142 or 242 can also be based on The block position of process object, judges whether upper piece control parameter can be utilized in decoding, can not utilize upper piece control In the case of parameter, use pattern 2 or pattern 3 determine context.And then, context control unit 142 or 242 can also be based on The tree construction of TU, CU or PU block segmentation decides whether that the reference point of block on this can be utilized.That is, context control unit 142 or 242 Whether the depth of the level of data unit that can also be according to belonging to the control parameter of process object judges being capable of profit in decoding With piece control parameter.
Fig. 9 B are the figures for the relation for representing picture (picture) in HEVC standard, section and block.1 picture is divided For more than 1 section.In the example shown in Fig. 9 B, picture is divided into two sections (SLICE1 and SLICE2).1 is cut Piece is made up of multiple pieces 301 (such as treeblocks).Here, in the case of the size as defined in section is divided into of block 301, It is maximum unit as certain control unit, is the root in the case of the root (root) for splitting the unit as level Size.
In the example shown in Fig. 9 B, SLICE2 is by since block 301A (startTb) including via by the block of painting shade 301 arrive multiple pieces of 1 Sequence composition of the block in the lower right corner.In addition, by the block of painting shade being as current processing pair in figure 1 block (Treeblock) of elephant.
Each block 301 is made up of N × M pixels.In addition, 1 block 301 (is typically internally 4 points by recurrence Ground Split Cut).That is, 1 Treeblock conceptually constitutes 14 fork tree (QuadTree).In the block 301B shown in Fig. 9 B, 4 segmentations The block of upper right afterwards is throughout 2 levels by recursively 4 segmentations.That is, No. 9 from No. 0 of upper left to bottom right of block 301B, including to provide Viewpoint segmentation 10 logic units.
Here, so-called viewpoint, refers to Tree on the tree of coding units (CU) or on residual_data etc., Using some root as the possible different multiple trees of the mutual depth of basic point.Here, the value of various control parameters belongs to some leaf node.
Here, whether the value of the control parameter of actual some signal kinds included in upper piece " can utilize (available) species for the tree that the control parameter belongs to ", is depended on.Thus, context control unit 142 or 242 can also be by Species according to the tree belonging to control parameter to become further determines specification.This is of equal value with being changed to syntactical unit.For example, context Control unit 142 or 242 can also be used the data of the alf_param on sef-adapting filter etc. data configuration and do not utilized Upper piece pattern 2 or pattern 3, context model policy (pattern 1) as prior art is used for other grammers.That is, on It can also be the control parameter with data configuration set in advance to state the 2nd species or the 3rd species.In addition, this means basis The species of the tree of adjacent definition and change.
It is actual whether to utilize the value of the control parameter or whether be imitated in the reduction of memory usage amount Really, it is different according to the position in the hierarchical relationship of block.That is, context control unit 142 or 242 can also according to the level of block and Whether the position switching in level utilizes upper piece.
For example, it is assumed that each number 0~9 of the block 301B shown in Fig. 9 B is decoding order.In the case, by number 4 Block coding or decoding in the case of, can using number 1 block and number 2 block control parameter.
And then, from the viewpoint of the reduction of memory usage amount, context control unit 142 or 242 can also be if be Depth0 block and be the position of oneself for block in the longitudinally split key element of more than second then Selection utilization pattern 1. Here, so-called depth (depth), represents the number of levels from root.That is, some slip gauge is being set to block [xn], [y0] Whether can also be block [xn] [(y0)+1] [depth] according to the block of process object in the case of [depth]) set up block Specification is further determined to become.That is, upper piece is utilized to the block of the number 4~9 shown in Fig. 9 B.Because, if the coding of the tree or Decoding is the numerical order (since 0, with 9 orders terminated) of diagram, then obviously can utilize upper in the block of number 4~9 The control parameter included in block.And then, in these blocks, also data remain temporarily just can with advantage.In addition, this It can be described as determining context value always according to the three-dimensional position including level in addition to x, y-coordinate.Furthermore it is possible to use On (following) condition value of the block of level as the block of lower level condition value.
And then, context control unit 142 or 242 is it is also contemplated that the block of process object and the position relationship of other sections come Change these specifications.Hereinafter, 3 shown in Fig. 9 B block 301A, 301B and 301C for applying shade example is illustrated.
Here, block 301A be start block, left piece and upper piece be included in other section in.Block 301B upper piece be included in it is other Section in.Block 301C upper piece and left piece be included in in the section of the section identical of the block comprising itself.Context control Portion 142 or 242 processed can also switch specification according to such condition.That is, context control unit 142 or 242 (1) can also root Whether be included in other sections according to upper piece and switch specification, (2) can also according to left piece whether be included in other sections come Switch specification, (3) can also switch specification according to their both sides.In other words, context control unit 142 or 242 can also be In the case that the block of process object belongs to slice boundaries, it is judged as that in decoding upper piece control parameter can not be utilized.Thus, example In the case that decoding process such as section 1 up is not over, it can realize whether certainly to consider in the inside of section 2 Oneself can obtain the decoding process of information.
Hereinafter, hierarchical processing unit (multi-layer block construction) is illustrated.Figure 10 is hierarchical for illustrating Handle the explanation figure of unit (multi-layer block construction).
Above-mentioned picture coding device 100 is by moving image according to each processing unit encoding, and picture decoding apparatus 200 will be compiled Code stream is according to each processing unit decoding.Using by the processing unit as multiple less processing units, by the less place Reason unit is divided into the mode of multiple smaller processing units again, carries out hierarchical.In addition, processing unit is smaller, the processing list Level residing for position is deeper, in low layer, represents that the value of the level is larger.On the contrary, processing unit is bigger, residing for the processing unit Level it is more shallow, in high level, represent that the value of the level is smaller.
In processing unit, there are coding units (CU), prediction unit (PU) and change of scale (TU).CU is maximum by 128 The block that × 128 pixels are constituted, is comparable to the unit of the macro block of prior art.PU is the base unit of inter-picture prediction.TU is The base unit of orthogonal transformation, the size of the TU is the size of a rank identical with PU or smaller than PU.CU is for example divided into (in the case, PU and TU is in mutual including the PU and TU with the sub- CU identicals size by 4 sub- CU, 1 sub- CU therein Overlapping state).Also it is further segmented as 4 sub- TU for 4 sub- PU, TU for example, the PU is further segmented.In addition, In the case that processing unit is divided into multiple less processing units, the less processing unit is referred to as subprocessing unit. For example, in the case where processing unit is CU, subprocessing unit is sub- CU, and in the case where processing unit is PU, subprocessing list Position is sub- PU, and in the case where processing unit is TU, subprocessing unit is sub- TU.
Specifically, it is following such.
Picture is divided into section.Section is the sequence of maximum coding units.Compiled by maximum the position of maximum coding units Code unit address lcuAddr is represented.
Each coding units including maximum coding units are divided into 4 coding units.As a result, coding units are constituted The Quadtree Partition of size.The position of coding units is by the sample (pixel or coefficient) with the left upper end of maximum coding units The coding units index cuIdx of point is represented.
In the case of the segmentation of license coding units, the coding units are regard as prediction unit disposal.With coding units Equally, the position of prediction unit is by the prediction Unit Index puIdx tables using the sample of the left upper end of maximum coding units as starting point Show.
Predict that unit can also include multiple subregions (prediction unit subregion or sub- PU).Prediction unit subregion is by single to predict The sample of the left upper end of position represents for the prediction unit subregion index puPartIdx of starting point.
Prediction unit can also include multiple change of scale.Same with coding units, change of scale can also be divided into 4 The change of scale (sub- change of scale) of individual less size.The Quadtree Partition of this license residual signals.The position of change of scale Represented by the sample of the left upper end to predict unit as the change of scale index tuIdx of starting point.
Here, the definition of each processing unit is following such.
CTB (coding tree block encode tree block)：For determine square area Quadtree Partition it is basic Unit.CTB has the diversified size of square.
LCTB (largest coding tree block, maximum coding tree block)：The maximum size permitted in a slice CTB.Section is made up of unduplicated multiple LCTB.
SCTB (smallest coding tree block, minimum code tree block)：The minimum chi permitted in a slice Very little CTB.Disapprove and SCTB is divided into smaller CTB.
PU (prediction unit predict unit)：For determining the base unit of prediction processing.PU size with not The size for being licensed the CU of segmentation is identical.In CU, CU is divided into 4 square areas by license, in contrast, in PU, PU can be divided into multiple subregions of arbitrary shape.
TU (transform unit, change of scale)：For the base unit for determining to convert and quantifying.
CU (coding unit, coding units)：It is identical with CTB.
LCU (largest coding unit, maximum coding units)：It is identical with maximum CTB.
SCU (smallest coding unit, minimum code unit)：It is identical with minimum CTB.
In addition, quantization parameter include Delta quantization zooming parameter (delta QP or QP delta), quantization offset parameter, Index at least one in (Qmatrix select idc) and quantization dead band offset parameter.In addition, index is for from multiple amounts Change selection 1 in scaled matrix.
Delta quantization zooming parameter (delta QP or QP delta) is the quantization zooming that should be applied in conversion coefficient Parameter, with by sequence head or section head specify quantization zooming parameter (or in z scanning sequencies it is back to back before quantization contract Put parameter) difference.
Quantization offset parameter is also referred to as quantifying skew, is the adjusted value (skew for rounding mode of signal when being quantified Value).Thus, picture coding device 100 is quantified skew coding when being quantified, and picture decoding apparatus 200 encodes this Quantization skew decoding afterwards.Also, picture decoding apparatus 200 is when by conversion coefficient re-quantization, progress is offset using the quantization Amendment.
Index (Qmatrix select idc) is also referred to as adaptive quantizing matrix, is represented from multiple quantization zooming matrixes The middle index using which quantization zooming matrix.In addition, feelings of the Qmatrix select idc in only 1 quantization zooming matrix Indicate whether to use the quantization zooming matrix under condition.In addition, adaptive application amount matrix can be controlled with block unit (processing unit) System.
Quantify dead band offset parameter and be also referred to as adaptive dead zone, for dead band is changed according to each block adaptive Control information.Dead band is coefficient of frequency by being quantized into the amplitude (being after quantization the amplitude before+1 or -1) for 0.
In addition, in the above description, describing as pattern 3 and using fixed value set in advance as context value Example, but as long as be to upper piece and left piece of control parameter can be can also be used not as pattern 3 not the condition utilized Comprising upper piece and left piece of control parameter condition condition.For example, as pattern 3, context control unit 142 or 242 also may be used Context is determined with the depth of the level of the data unit according to belonging to control parameter.
(embodiment 3)
In the present embodiment, to what kind of signal should be utilized as above-mentioned 1st species and the 2nd species (or the 3rd species) Species is illustrated.
Specifically, present inventor is in the signal kinds shown in Fig. 3 (non-patent literature 2,9.3.3.1.1.1 sections) Each following signal kinds are verified.It is because parameter is throughout multiple branches, root that each signal kinds are carried out with checking Each mould of other signal kinds is difficult to according to the result (1~pattern of pattern 3 which more preferable) on 1 signal kinds Whether formula meets properness.
Checking foundation is in JCTVC-E700, " Common test conditions and software reference Structure (setup parameter, and software version HM3.0) described in configurations " (with reference to non-patent literature 3).In addition, It is that the length of test image is all restricted to the result of 49 frames.
Method for encoding images and picture decoding method about present embodiment are on CABAC.Thus, as Entropy Coding (entropy code) pattern, is 1 (#0 using the value as SymbolMode：LCEC, 1：CABAC setting value) Following 4 test pattern of group verified.
4.1Intra, high-efficiency setting
4.3Random access, high-efficiency setting
4.5Low delay, high-efficiency setting
4.7Low delay, high-efficiency setting (P slices only)
In addition, evaluating the referred to as " BD- for using the installation in HEVC to be utilized in evaluating as unified opinion scale The evaluation of estimate of RATE " values.Y BD-rate, U BD-rate and V BD-rate are the BD-rate on color space Y, U, V, are Opinion scale value.Here, so-called BD-rate, be two groups of size of codes will being documented in VCEG-AI11 (non-patent literature 4) and PSNR result integration, by the ratio of its area come the evaluation of estimate of presentation code efficiency.In addition, representing negative value in BD-rate In the case of presentation code efficiency improve.The benchmark compared is the output result of the reference program of Installation Modes 1, this result table Show the respective value relative to pattern 1 of pattern 2 and pattern 3.
Hereinafter, the result of each signal kinds is illustrated.
(the 1st checking) " split_coding_unit_flag "
(the 2nd checking) " skip_flag "
(the 3rd checking) " merge_flag "
(the 4th checking) " ref_idx_l0 (l1, lc) "
(the 5th checking) " inter_pred_flag "
(the 6th checking) " mvd_l0 (l1, lc) "
(the 7th checking) " no_residual_data_flag "
(the 8th checking) " intra_chroma_pred_mode "
(the 9th checking) " cbf_luma, cbf_cr, cbf_cb "
(the 1st checking) " split_coding_unit_flag "
Figure 11 is the table illustrated for the arithmetic decoding method to split_coding_unit_flag.
In the verification, only context model slave pattern 1 is changed to pattern 2 or pattern 3 by the signal kinds of identifying object, is closed Verified in other signal kinds and the checking specified by non-patent literature 3 with parameter without change.In addition, In value in Figure 11 column, so-called " fixation " is without using recording " fixation " in the export of context value (or increment size) Row condition (without using left piece of condition or upper piece condition).That is, it is in the condition of only left piece condition and a side of right piece of condition In the case of " fixation ", using only the condition of the opposing party.In addition, left piece of condition and right piece of condition both be " fixation " feelings Under condition, value set in advance (such as " 0 ") is used as context value (or increment size).
The meaning of signal kinds " split_coding_unit_flag " is defined below.
Split_coding_unit_flag [x0] [y0] represents whether the CU of object is divided into the chi of horizontal and vertical Very little multiple CU for half.X0 and y0 represent CU on the basis of the luma samples positioned at the upper left corner of image, positioned at object The upper left corner luma samples coordinate.That is, whether split_coding_unit_flag represents the CU of object by 4 segmentations.Tool For body, in the case where split_coding_unit_flag value is " 1 ", the CU of object is divided, and is " 0 " in the value In the case of, the CU of object is not divided.
As syntax, data configuration turns to Coding treesyntax to the split_coding_unit_flag.Figure Grammar explanation bit sequence as decoding apparatus according to the data configuration.
Figure 12 A and Figure 12 B are the tables for representing the result on split_coding_unit_flag.
Figure 12 A represent the result in the case of use (pattern 2) 1 adjacent block (only left piece condition L judgment value). Figure 12 B represent to use the result in the case of (upper block condition L and a left side are not all used) of (pattern 3) 0 adjacent block.
In addition, the result shown in Figure 12 A and Figure 12 B represents the increase and decrease of the BD-Rate values of above-mentioned 4 test patterns.
In addition, evaluation of estimate is to represent relative with the evaluation of estimate using left piece and in the case of upper piece both patterns 1 Value opinion scale.Specifically, if evaluation of estimate is on the occasion of its result is than the evaluation of estimate (BD- in the case of pattern 1 Rate it is) poor.In addition, if evaluation of estimate is negative value, the evaluation of estimate in the case that its result compares pattern 1 is enhanced.
It can be seen from the result, on split_coding_unit_flag, the pattern of context model, pattern 1 are used as Preferably.In other words, the evaluation of estimate of pattern 2 and pattern 3 is poorer than pattern 1.
Thus, context control unit 142 and 242 is " split_coding_unit_ in the signal kinds of control parameter In the case of flag ", from BD-rate viewpoint, the pattern 1 by the use of the pattern as context model as prior art is determined Determine context value.
(the 2nd checking) " skip_flag "
Figure 13 is the table illustrated for the arithmetic decoding method to skip_flag.In addition, verification method and above-mentioned the 1 checking is same.
The meaning of signal kinds " skip_flag " is defined below.
In the case of by P or B slice decoders, in the case where skip_flag [x0] [y0] value is " 1 ", it is meant that After the skip_flag [x0] [y0], in addition to predicted motion vector index, in the absence of current CU grammatical feature. In addition, in the case where skip_flag [x0] [y0] value is " 1 ", it is meant that current CU can not be skipped.X0 and y0 are represented The coordinate of the luma samples in CU on the basis of the luma samples positioned at the upper left corner of image, positioned at object the upper left corner.That is, Skip_flag indicates whether to skip the CU of object and (whether be used as the block disposal being skipped).
Coding unit syntax are arrived by data configurationization as syntax, the skip_flag.That is, skip_flag quilts Set according to each CU.In picture decoding apparatus, the grammar explanation bit sequence according to the data configuration.
Figure 14 A and Figure 14 B are the tables for representing the result on skip_flag.
Figure 14 A represent the result in the case of use (pattern 2) 1 adjacent block (only left piece condition L judgment value). Figure 14 B represent to use the result in the case of (pattern 3) 0 adjacent block (upper block condition L and left without using).
In addition, the result and the 1st checking shown in Figure 14 A and Figure 14 B are same, the BD- of above-mentioned 4 test patterns is represented The increase and decrease of Rate values.In addition, the meaning of evaluation of estimate is same with the 1st checking.
It can be seen from the result, on " skip_flag ", as the pattern of context model, pattern 1 is preferable.In other words, The evaluation of estimate of pattern 2 and pattern 3 is poorer than pattern 1.
Thus, context control unit 142 and 242 is in the case where the signal kinds of control parameter are " skip_flag ", from BD-rate viewpoint, the pattern 1 by the use of the pattern as context model as prior art determines context value.
(the 3rd checking) " merge_flag
Figure 15 is the table illustrated for the arithmetic decoding method to merge_flag.In addition, verification method with it is above-mentioned 1st checking and the 2nd checking are identical.
The meaning of signal kinds " merge_flag " is defined below.
Merge_flag [x0] [y0] represents whether current CU inter prediction parameter is pre- through interframe by closing on The block export measured.X0 and y0 represent prediction block on the basis of the luma samples positioned at the upper left corner of image, positioned at object The upper left corner luma samples coordinate.In the absence of merge_flag [x0] [y0] situation (InferredMergeFlag value It is the situation of " 1 ") under, merge_flag [x0] [y0] is " 1 " by inference.That is, merge_flag [x0] [y0] indicates whether to make Use merging patterns.Here, merging patterns refer to replicate motion vector and reference picture index from the adjacent block of coded object block, and Carry out the pattern of the coding of coded object block.
" Prediction Unit " is arrived by data configurationization as syntax, the merge_flag.That is, merge_flag It is set according to each PU.In picture decoding apparatus, the grammar explanation bit sequence according to the data configuration.
Figure 16 A and Figure 16 B are the tables for representing the result on merge_flag.
Figure 16 A represent the result of study in the case of use (pattern 2) 1 adjacent block (only left piece condition L judgment value). Figure 16 B represent to use the result in the case of (pattern 3) 0 adjacent block (upper block condition L and left without using).
In addition, the result and the 1st checking shown in Figure 16 A and Figure 16 B are same, the BD- of above-mentioned 4 test patterns is represented The increase and decrease of Rate values.In addition, the meaning of evaluation of estimate is same with the 1st checking.
According to the result, as a result with the skip_ of the checkings of split_coding_unit_flag and the 2nd of the above-mentioned 1st checking Flag is different, on the merge_flag, is not sent out between pattern 1 and pattern 2 or pattern 3 as the pattern of context model Raw BD-rate larger difference.
Thus, context control unit 142 and 242 is special in the environment of control parameter of multiple signal kinds is mixed Not in the case where signal kinds are merge_flag, determine as adjacent block without utilizing upper piece context value.I.e., up and down Literary control unit 142 and 242 is in the case where the signal kinds of control parameter are merge_flag, and use pattern 2 or pattern 3 are determined Context value.In other words, above-mentioned 1st species includes " split_coding_unit_flag " or " skip_flag ", and the above-mentioned 2nd Species or the 3rd species include " merge_flag ".Thus, about the picture coding device and picture decoding apparatus of present embodiment Memory usage amount can be cut down while the decline of BD-rate values is suppressed.
In addition, in the case where the pattern 2 for merge_flag is compared with pattern 3, in their BD-rate not Occurs larger difference.Use pattern 3 is more preferably accordingly, for merge_flag.Thereby, it is possible to realize further storage The reduction of device usage amount and the reduction for the treatment of capacity.
Here, merge_flag and skip_flag are compared, there is following difference：Motion vector is not sent when skipping Residual error data, but in merging patterns send motion vector residual error data.This reason it is assumed that merge_flag is used Context is not that when being best suitable for, by using the processing of the residual error data, can make up to a certain extent due to that can not make yet With the decline of image quality caused by most suitable context.In this way it can be considered that the decline of image quality can be suppressed.
(the 4th checking) " ref_idx_l0 (l1, lc) "
Figure 17 is the table illustrated for the arithmetic decoding method to ref_idx.In addition, verification method and above-mentioned the 1 checking and the 2nd checking are identical.
The meaning of signal kinds " ref_idx " is defined below.
Ref_idx_l0 [x0] [y0] represents the reference picture index of current PU list 0.X0 and y0 are represented with position The coordinate of the luma samples in the upper left corner of prediction block on the basis of the luma samples in the upper left corner of image, positioned at object.
In addition, ref_idx_l0 [x0] [y0] has with the l0 in refref_idx_l0 and list 0 are replaced into l1 respectively And identical meaning during list 1.That is, ref_idx_l1 represents the reference picture index of current PU list 1.
The presence or absence of ref_idx_l1 can be determined according to picture categories etc..
In addition, ref_idx_lc [x0] [y0] has with the l0 in refref_idx_l0 and list 0 are replaced into lc respectively And the same meaning of synthesis list phase.The ref_idx_lc is additional control parameter in HAVC.Also, synthesis list be by List 0 and list 1 synthesize the list after (merging).Generally, in ref_idx_l0 and ref_idx_l1 and ref_idx_lc only One side is comprising in the bitstream.In addition, both there is ref_idx_l0 and ref_idx_l1 an only side includes feelings in the bitstream Condition, situation in the bitstream is included there is also both sides.
" Prediction Unit " is arrived by data configurationization as syntax, the ref_idx_l0 (l1, lc).In image During decoding makeup is put, bit sequence is explained according to the grammer of the data configuration.
Figure 18 A and Figure 18 B are the tables for representing the result on ref_idx.
Figure 18 A represent the result in the case of use (pattern 2) 1 adjacent block (only left piece condition L judgment value). Figure 18 B represent the result in the case of use (pattern 3) 0 adjacent block (upper block condition L and a left side are all not used).
In addition, the result and the 1st checking shown in Figure 18 A and Figure 18 B are same, the BD- of above-mentioned 4 test patterns is represented The increase and decrease of Rate values.In addition, the meaning of evaluation of estimate is same with the 1st checking.
According to the result, as a result with the skip_ of the checkings of split_coding_unit_flag and the 2nd of the above-mentioned 1st checking Flag is different, on the ref_idx, does not occur between pattern 1 and pattern 2 or pattern 3 as the pattern of context model BD-rate larger difference.
Thus, context control unit 142 and 242 is special in the environment of control parameter of multiple signal kinds is mixed It is unfavorable with piece determining context value as adjacent block not in the case where signal kinds are ref_idx_l0 (l1, lc). That is, context control unit 142 and 242 uses mould in the case where the signal kinds of control parameter are ref_idx_l0 (l1, lc) Formula 2 or pattern 3 determine context value.In other words, above-mentioned 1st species includes " split_coding_unit_flag " or " skip_ Flag ", above-mentioned 2nd species or the 3rd species include " ref_idx_l0 (l1, lc) ".Thus, compiled about the image of present embodiment Code device and picture decoding apparatus can cut down memory usage amount while the decline of BD-rate values is suppressed.
As long as in addition, above-mentioned 2nd species or the 3rd species are including in ref_idx_l0, ref_idx_l1 and ref_idx_lc At least one.For example, above-mentioned 2nd species or the 3rd species can include ref_idx_l0 and ref_idx_l1, without including ref_idx_lc。
In addition, in the case where comparing pattern 2 and pattern 3 for ref_idx, not occurring in their BD-rate Larger difference.Use pattern 3 is more preferably accordingly, for ref_idx_l0 (l1, lc).It is further thereby, it is possible to realize The reduction of memory usage amount and the reduction for the treatment of capacity.
Here, used in the case where not being to skip common interframe (inter) predictive mode of pattern and merging patterns ref_idx.In skip mode and merging patterns, as process object block motion vector be utilize with upper piece and left piece it is identical Motion vector, be to utilize as the motion vector of process object block but in common inter-frame forecast mode in addition From upper piece and left piece of different motion vector.Thus, if utilizing upper and left piece as pattern 1 to ref_idx, it can recognize Decline for code efficiency.That is, by that to ref_idx use patterns 2 or pattern 3, code efficiency can be improved.
Here, context control unit 142 and 242 is judging condition A (or the conditions on ref_idx_lc [xP] [yP] When L), the result and condA result of the above-mentioned condL from the ref_idx_l0 [xP] [yP] on current block can also be used Derived value, and without using adjacent block A or B ref_idx_lc [xL] [yP] or ref_idx_lc [xP] [yA].Change speech It, the end value that context control unit 142 and 242 can also export condition is used as the subordinate value of l0, l1 of current block value.
Code device or tape deck by when carrying out stream record or during coding by ref_idx_l0 and ref_ Idx_l1 is synthesized, and generates ref_idx_lc.That is, all using ref_idx_l0 and ref_ in the judgement of these devices internally idx_l1.Therefore, in current block, (1) condition condA or L for (ref_idx_l0 is available) and (ref_idx_l0>0 is true set up) } and (2) condition condA or L for (ref_idx_l1 is available) and (ref_idx_l1>0 is true set up) } in the case of, condition condA or L for (ref_idx_lc is available) and (ref_idx_lc>0 is true set up) } also set up.
Accordingly it is also possible to be discussed further below.Figure 18 C are the condition A and bar for representing ref_idx disclosed in non-patent literature 2 Part L table.Figure 18 D are the tables for representing condition A and condition L about the ref_idx of present embodiment.
As shown in Figure 18 D, context control unit 142 and 242 can also be from the ref_idx_l0 and ref_idx_ of same At least one party exports ref_idx_lc condition value condL and condA in l1 condition value.That is, context control unit 142 and 242 can also make ref_idx_lc condition value condL and condA linear dependent in the ref_idx_l0 and ref_ of same Idx_l1 condition value.
This way it is not necessary to the memory reference on ref_idx_lc.That is, can be without reference to upper piece ref_idx_lc Value in the case of export ref_idx_lc condition value.
(the 5th checking) " inter_pred_flag "
Figure 19 is the table illustrated for the arithmetic decoding method to inter_pred_flag.
The meaning of signal kinds " inter_pred_flag " is defined below.
Inter_pred_flag [x0] [y0] according to table 7 11, represent to current PU using uni-prediction and Bi-prediction which.X0 and y0 represent on the basis of the luma samples positioned at the upper left corner of image, positioned at object The coordinate of the luma samples in the upper left corner of prediction block.Here, uni-prediction is the prediction using lc (synthesis list), Bi-prediction is the prediction using list 0 and list 1.In addition, synthesis list is to synthesize list 0 and list 1 (merging) List afterwards.In addition, inter_pred_flag is only used in the case where object slice is B sections.
" Prediction Unit " is arrived by data configurationization as syntax, the inter_pred_flag.In image solution During code makeup is put, the grammar explanation bit sequence according to the data configuration.
Figure 20 A and Figure 20 B are the tables for representing the result on inter_pred_flag.
Figure 20 A represent the result in the case of use (pattern 2) 1 adjacent block (only left piece condition L judgment value). Figure 20 B represent the result in the case of use (pattern 3) 0 adjacent block (upper block condition L and a left side are all not used).
In addition, the result and the 1st checking shown in Figure 20 A and Figure 20 B are same, the BD- of above-mentioned 4 test patterns is represented The increase and decrease of Rate values.In addition, the meaning of evaluation of estimate is same with the 1st checking.
According to the result, as a result with the skip_ of the checkings of split_coding_unit_flag and the 2nd of the above-mentioned 1st checking Flag is different, on the inter_pred_flag, as the pattern of context model pattern 1 and pattern 2 or pattern 3 it Between BD-rate larger difference does not occur.
Thus, context control unit 142 and 242 is special in the environment of control parameter of multiple signal kinds is mixed It is unfavorable with piece determining context value as adjacent block not in the case where signal kinds are inter_pred_flag.That is, Context control unit 142 and 242 in the case where the signal kinds of control parameter are inter_pred_flag, use pattern 2 or Pattern 3 determines context value.In other words, above-mentioned 1st species includes " split_coding_unit_flag " or " skip_ Flag ", above-mentioned 2nd species or the 3rd species include " inter_pred_flag ".Thus, about the Image Coding of present embodiment Device and picture decoding apparatus can cut down memory usage amount while the decline of BD-rate values is suppressed.
In addition, in the case where comparing pattern 2 and pattern 3 for inter_pred_flag, their BD-rate In larger difference does not occur.Use pattern 3 is more preferably accordingly, for inter_pred_flag.Thereby, it is possible to realize into The reduction and the reduction for the treatment of capacity of the memory usage amount of one step.
Here, used in the case where not being to skip common interframe (inter) predictive mode of pattern and merging patterns inter_pred_flag.In skip mode and merging patterns, as process object block motion vector and utilize with upper piece and Left piece of identical motion vector, but in common inter-frame forecast mode in addition, sweared as the motion of process object block Amount and utilize from upper piece and left piece of different motion vector.Thus, if utilized to inter_pred_flag as pattern 1 Go up and left piece, then it is considered that code efficiency declines.That is, can by inter_pred_flag use patterns 2 or pattern 3 Improve code efficiency.
And then, as described above, for inter_pred_flag, above and below being determined according to the level of process object block Text value, can further improve code efficiency.
(the 6th checking) " mvd_l0 (l1, lc) "
Figure 21 is the table illustrated for the arithmetic decoding method to mvd_l0 (l1, lc).
Signal kinds mvd_l0 (l1, lc) meaning is defined below.
Mvd_l0 [x0] [y0] [compIdx] represents the vector components and the difference of its predicted value of used list 0.x0 And y0 represents the brightness sample in the upper left corner of prediction block on the basis of the luma samples positioned at the upper left corner of image, positioned at object This coordinate.Difference to the motion vector composition of horizontal direction distributes compIdx=0.Also, the motion to vertical direction is sweared Measure composition distribution compIdx=1.In the case of two compositions are all non-existent, value is considered as [0].That is, mvd_l0 is to use 1 composition (horizontal component compIdx=0) and the 2nd composition (vertical component compIdx=1) state some PU position (xP, yP) The difference of motion vector and predictive vector.
Mvd_l1 [x0] [y0] [compIdx] has with the l0 in mvd_l0 and list 0 are replaced into l1 and list 1 respectively When identical meaning.The presence or absence of the mvd_l1 can be determined according to picture categories etc..
In addition, mvd_lc [x0] [y0] [compIdx] have with by the l0 in mvd_l0 and list 0 be replaced into respectively lc and Synthesize the same meaning of list phase.That is, mvd_lc is generated by the way that mvd_l0 and mvd_l1 is synthesized.
Below, in referred to as mvd, the mvd at least include mvd_l0, and according to the condition of image include mvd_l1 and At least one party in mvd_lc.
" Prediction Unit " is arrived by data configurationization as syntax, the mvd.In image decoding makeup is put, according to According to the grammar explanation bit sequence of the data configuration.
Figure 22 A and Figure 22 B are the tables for representing the result on mvd.
Figure 22 A represent the result in the case of use (pattern 2) 1 adjacent block (only left piece condition L judgment value). Figure 22 B represent the result in the case of use (pattern 3) 0 adjacent block (upper block condition L and a left side are all not used).
In addition, the result and the 1st checking shown in Figure 22 A and Figure 22 B are same, the BD- of above-mentioned 4 test patterns is represented The increase and decrease of Rate values.In addition, the meaning of evaluation of estimate is same with the 1st checking.
According to the result, as a result with the skip_ of the checkings of split_coding_unit_flag and the 2nd of the above-mentioned 1st checking Flag is different, on the mvd, does not occur BD- between pattern 1 and pattern 2 or pattern 3 as the pattern of context model Rate larger difference.
Thus, context control unit 142 and 242 is special in the environment of control parameter of multiple signal kinds is mixed It is unfavorable with piece determining context value as adjacent block not in the case where signal kinds are mvd_l0 (l1, lc).That is, on Hereafter control unit 142 and 242 is in the case where the signal kinds of control parameter are mvd_l0 (l1, lc), use pattern 2 or pattern 3 determine context value.In other words, above-mentioned 1st species includes " split_coding_unit_flag " or " skip_flag ", on Stating the 2nd species or the 3rd species includes " mvd_l0 ", " mvd_l1 " or " mvd_lc ".Thus, about the image of present embodiment Code device and picture decoding apparatus can cut down memory usage amount while the decline of BD-rate values is suppressed.
As long as in addition, above-mentioned 2nd species or the 3rd species include at least one in mvd_l0, mvd_l1 and mvd_lc. For example, above-mentioned 2nd species or the 3rd species can include mvd_l0 and mvd_l1, without including mvd_lc.
In addition, in the case where comparing pattern 2 and pattern 3 for mvd, not occurring in their BD-rate larger Difference.Use pattern 3 is more preferably accordingly, for mvd_l0 (l1, l0).Thereby, it is possible to realize that further memory makes The reduction of consumption and the reduction for the treatment of capacity.
Here, do not send the residual error data (mvd) of motion vector when skipping, but motion arrow is sent in merging patterns The residual error data (mvd) of amount.This reason it is assumed that the context used in merging patterns is not most suitable, but by using this Mvd processing, the decline of the image quality caused by it can not use most suitable context can be also made up to a certain extent. Based on this reason, it is believed that suppress the decline of image quality in the case where not utilizing the block on periphery for mvd.
According to foregoing defined condition, in utilization upper piece or the situation of left piece of condition value (condA or condL) Under, it can also be applicable following deformation.
First variation is the method for the subordinate relation using mvd_l0 and mvd_l1 and mvd_lc.
Specifically or, context control unit 142 and 242 use mvd_l0 and mvd_l1 and mvd_lc this 3 The condition value (condL or condA) of 2 signal kinds in signal kinds, export, which has, is subordinated to the another of their condition value The condition value of one signal kinds.
For example, in condition value (the l0 condA value or l1 relative to the two signal kinds of mvd_l0 and mvd_l1 CondA value), in the case that mvd_lc condA value is subordinate relation, context control unit 142 and 242 need not join According to the value of the condA on mvd_lc.
Figure 22 C are the condition A and condition L that represent mvd disclosed in non-patent literature 2 table.Figure 22 D are to represent this relevant reality Apply the mvd of mode condition A and condition L table.
As shown in figure 22d, context control unit 142 and 242 can also be from the mvd_l0 and mvd_l1 of same condition At least one party exports mvd_lc condition value condL and condA in value.
In addition, their relation can also be applied to horizontal direction (compIdx=0) by context control unit 142 and 242 Or one side or both sides in vertical direction (compIdx=1).
In addition, context control unit 142 and 242 can also utilize the subordinate relation between compIdx=0 and 1.That is, on Hereafter control unit 142 and 242 can also make horizontal direction mvd_l0 [] [] [0] and vertical direction mvd_l0 [] [] [1] this two The result of side's condition value is subordinated to the opposing party in individual condition value.That is, context control unit 142 and 242 can also be from the opposing party The condition value condL and condA of a side mvd in mvd condition value lead-out level direction and vertical direction.Here, in non-patent In document 2, mvd (mvd_l0 [] [] [0], mvd_l1 [] [] [0], mvd_lc [] [] [0]) for horizontal direction and vertical The mvd (mvd_l0 [] [] [1], mvd_l1 [] [] [1], mvd_lc [] [] [1]) in direction sets context index (rope respectively Draw increment size+a reference value).By using above-mentioned subordinate relation, this waste can be cut down.That is, context rope can be cut down The quantity drawn.
Here, in the record of non-patent literature 2, upper block and left piece of condition have been used only for mvd the 1st bit Value.That is, context control unit 142 and 242 can also use above-mentioned pattern 2 or pattern 3 for mvd the 1st bit.Change speech It, whether context control unit 142 and 242 can also be more than 0 for the difference for representing motion vector and predictive vector Abs_mvd_greater0_flag [compIdx] uses above-mentioned pattern 2 or pattern 3.
(the 7th checking) " no_residual_data_flag "
Figure 23 A are the tables illustrated for the arithmetic decoding method to " no_residual_data_flag ".
The meaning of signal kinds " no_residual_data_flag " is defined below.
In the case where no_residual_data_flag is " 1 ", in the absence of current CU residual error data.In no_ In the case that residual_data_flag is " 0 ", there is current CU residual error data.In no_residual_data_ In the case of flag is non-existent, its value is considered as " 0 ".
As syntax, the no_residual_data_flag is by among the species of data configurationization to above-mentioned tree In Transform Tree.Figure 23 B are the tables for representing Transform Tree syntax.In image decoding makeup is put, foundation The grammar explanation bit sequence of the data configuration.
Figure 24 A and Figure 24 B are the tables for representing the result on " no_residual_data_flag ".
Figure 24 A represent the result in the case of use (pattern 2) 1 adjacent block (only left piece condition L judgment value). Figure 24 B represent the result in the case of use (pattern 3) 0 adjacent block (upper block condition L and a left side are all not used).
In addition, the result and the 1st checking shown in Figure 24 A and Figure 24 B are same, the BD- of above-mentioned 4 test patterns is represented The increase and decrease of Rate values.In addition, the meaning of evaluation of estimate is same with the 1st checking.
According to the result, as a result with the skip_ of the checkings of split_coding_unit_flag and the 2nd of the above-mentioned 1st checking Flag is different, on the no_residual_data_flag, as the pattern of context model in pattern 1 and pattern 2 or mould Do not occur BD-rate larger difference between formula 3.
Thus, context control unit 142 and 242 is special in the environment of control parameter of multiple signal kinds is mixed It is unfavorable with piece determining context as adjacent block not in the case where signal kinds are no_residual_data_flag Value.That is, context control unit 142 and 242 is no_residual_data_flag situation in the signal kinds of control parameter Under, use pattern 2 or pattern 3 determine context value.In other words, above-mentioned 1st species includes " split_coding_unit_ Flag " or " skip_flag ", above-mentioned 2nd species or the 3rd species include " no_residual_data_flag ".Thus, it is relevant The picture coding device and picture decoding apparatus of present embodiment can cut down storage while the decline of BD-rate values is suppressed Device usage amount.
In addition, in the case where comparing pattern 2 and pattern 3 for no_residual_data_flag, they Do not occur larger difference in BD-rate.Therefore, it is more preferably use pattern 3 for no_residual_data_flag.By This, can realize the reduction of further memory usage amount and the reduction for the treatment of capacity.
Here, no_residual_data_flag represents whether there is the coefficient (residual error) of brightness and aberration.Also, the residual error Increase in infra-frame prediction, reduce in inter prediction.Therefore, in the pattern of the block of the surrounding feelings different from the pattern of object block Under condition (in the case that characteristic is different), code efficiency declines.For example, object block be carried out infra-frame prediction, around block by reality In the case of applying inter prediction, the residual error of object block reduces, the residual error increase of the block of surrounding.Therefore, if upper around use Hereafter, then code efficiency declines.Thus, context control unit 142 and 242 can be improved by using the context of object block Code efficiency, and independent of the block of surrounding.
(the 8th checking) " intra_chroma_pred_mode "
Figure 25 A are the tables illustrated for the arithmetic decoding method to intra_chroma_pred_mode.
" PredictionUnit " is arrived by data configurationization as syntax, the intra_chroma_pred_mode. During image decoding makeup is put, the grammar explanation bit sequence according to the data configuration.
The meaning of signal kinds " intra_chroma_pred_mode " is defined below.
Intra_chroma_pred_mode [x0] [y0] represents the intra prediction mode of aberration sample.X0 and y0 are represented The seat of the luma samples in the upper left corner of prediction block on the basis of the luma samples positioned at the upper left corner of image, positioned at object Mark.
By the value (less than more than 04 value) and the IntraPredMode of current block of the intra_chroma_pred_mode [xP] [yB] combines, and determines that the face inner estimation mode value on aberration is " chroma intra prediction mode " (IntraPredModeC).Here, the coordinate of current block is set to [xB] [yB].It is somebody's turn to do [xB] [yB] and represents identical with [xP] [yP] Position.Also, IntraPredMode is the predictive mode value on brightness.
Figure 25 B be represent the intra_chroma_pred_mode described in non-patent literature 2 and with The table of the corresponding IntraPredModeC of IntraPredMode deriving method.
In addition, the IntraPredMode (variable of capitalization beginning) shown in Figure 25 A is not to be coded directly as compiling The value of code row, but the arrangement (array) of the value after decoding process is carried out inside decoder.In object block During IntraPredMode [xP] [yP] export processing, if it is possible to utilize IntraPredMode [xP] [yA] of adjacent block etc. Just it is used.
Figure 26 A and Figure 26 B are the tables for representing the result on " intra_chroma_pred_mode ".
Figure 26 A represent the result in the case of use (pattern 2) 1 adjacent block (only left piece condition L judgment value). Figure 26 B represent the result in the case of use (pattern 3) 0 adjacent block (upper block condition L and a left side are all not used).
In addition, the result and the 1st checking shown in Figure 26 A and Figure 26 B are same, the BD- of above-mentioned 4 test patterns is represented The increase and decrease of Rate values.In addition, the meaning of evaluation of estimate is same with the 1st checking.
According to the result, as a result with the skip_ of the checkings of split_coding_unit_flag and the 2nd of the above-mentioned 1st checking Flag is different, on the intra_chroma_pred_mode, as the pattern of context model in pattern 1 and pattern 2 or Do not occur BD-rate larger difference between pattern 3.
Thus, context control unit 142 and 242 is special in the environment of control parameter of multiple signal kinds is mixed It is unfavorable with piece determining context as adjacent block not in the case where signal kinds are intra_chroma_pred_mode Value.That is, context control unit 142 and 242 is intra_chroma_pred_mode situation in the signal kinds of control parameter Under, use pattern 2 or pattern 3 determine context value.In other words, above-mentioned 1st species includes " split_coding_unit_ Flag " or " skip_flag ", above-mentioned 2nd species or the 3rd species include " intra_chroma_pred_mode ".Thus, it is relevant The picture coding device and picture decoding apparatus of present embodiment can cut down storage while the decline of BD-rate values is suppressed Device usage amount.
In addition, in the case where comparing pattern 2 and pattern 3 for intra_chroma_pred_mode, they Do not occur larger difference in BD-rate.Therefore, it is more preferably use pattern 3 for intra_chroma_pred_mode.By This, can realize the reduction of further memory usage amount and the reduction for the treatment of capacity.
Here, intra_chroma_pred_mode is entirely the information of 4 bits, and 1 bit of beginning is represented for aberration Whether infra-frame prediction identical pattern with brightness is used.Here, 1 bit of the context control unit 142 and 242 for the beginning Use pattern 2 or pattern 3.Specifically, in the case where being used for aberration with brightness identical pattern, intra_ Chroma_pred_mode is " 0 ".In the case where using the pattern different from brightness for aberration, intra_chroma_ Pred_mode is " 1 ", and remaining 3 bit represents the pattern used for aberration.
Here, infra-frame prediction using upper piece and left piece it is related to object block.That is, due to having been utilized in infra-frame prediction Relevant information, thus utilize efficient with brightness identical model comparision for aberration.In other words, can also be utilized for aberration with The different pattern of brightness, to make pattern that there is diversity, but for aberration is using the pattern different from brightness and utilizes week The situation of the context enclosed is than sparser.That is, using with brightness identical pattern, intra_chroma_pred_mode set The situation for being set to " 0 " is in the majority.Therefore, the advantage of the context around use is less, can by using the context of object block Treating capacity is cut down while code efficiency is maintained.
In addition, it can be extremely difficult using " avaibale " to determine whether when decoding Dui upper piece.Reference picture 36 pairs export and the IntraPredMode for being arranged arrangement is illustrated in decoding process.
The IntraPredMode of the row (LineL) of upper 1 row of row including object block is represented as IntraPredMode [n-th of horizontal direction] [upper 1 row (LineL)].Also, the IntraPredMode of the current row including object block is expressed For IntraPredMode [k-th of horizontal direction] [current line].Here, it is as the signal of current arithmetic decoding object Intra_chroma_pred_mode [j-th of horizontal direction] [current line].
First, in the absence of above-mentioned upper 1 row row and current row n-th and k-th corresponding guarantor of horizontal direction Barrier.Because as described in Figure 36 explanation, the size of PU blocks changes because of each piece.Therefore, at desired pair Designed when they are managed certain corresponding table or as in Figure 36 illustrate as with least unit obtain it is all IntraPredMode, in addition without other methods.
In addition, different from the intra_chroma_pred_mode of decoder object, the IntraPredMode of row above is not It is the signal kinds that can be obtained by the analysis in arithmetic decoding portion, but the derived value by another decoding process (variable H.264 wait, capitalization beginning).Therefore, arithmetic decoding portion independently obtains the utilization possibility itself of the value Larger load will be produced.
Therefore, in context models of the intra_chroma_pred_mode using adjacent block, from memory usage amount Aspect says, it is unfavorable with piece condition value (especially with regard to the condition of IntraPredMode [correspondence horizontal level] [upper 1 row] Decision content) CondA is useful.
(the 9th checking) " cbf_luma, cbf_cr, cbf_cb "
Figure 27 is the table illustrated for the arithmetic decoding method to cbf_luma, cbf_cr and cbf_cb.
The meaning of signal kinds " cbf_luma " is defined below.
In the case where " cbf_luma [x0] [y0] [trafoDepth] " is " 1 ", luminance transformation block includes rank (level) be not 0 the conversion coefficient of more than 1.X0 and y0 are represented on the basis of the luma samples positioned at the upper left corner of image , the coordinates of the luma samples in the upper left corner of prediction block positioned at object.TrafoDepth is represented in the block of transition coding CU current segmentation rank.In the case where trafoDepth is " 0 ", block corresponds to CU.
That is, the position of luminance transformation block is defined to include key element vertically and horizontally and on the basis of CU level Relative hierarchical depth (trafoDepth) including three-dimensional arrangement value.
In addition, the meaning of signal kinds " cbf_cb " is used for providing the cbf_luma about brightness for aberration (Cb).Should The meaning of signal kinds " cbf_cb " is defined below.
In the case of cbf_cb [x0] [y0] [trafoDepth] " being " 1 ", Cb transform blocks include 1 that rank is not 0 Conversion coefficient above.X0 and y0 represent prediction on the basis of the luma samples positioned at the upper left corner of image, positioned at object The coordinate of the luma samples in the upper left corner of block.TrafoDepth represents the current segmentation level of the CU in the block of transition coding Not.In the case where trafoDepth is " 0 ", block corresponds to CU.Be not present in cbf_cb [x0] [y0] [trafoDepth] and And in the case that PredMode is different from MODE_INTRA, cbf_cb [x0] [y0] [trafoDepth] value is considered as " 0 ".
That is, the position of Cb transform blocks is defined to include key element vertically and horizontally and on the basis of CU level With respect to the value of the three-dimensional arrangement including hierarchical depth (trafoDepth).
In addition, the meaning of signal kinds " cbf_cr " is used for providing the cbf_luma about brightness for aberration (Cr).Should The meaning of signal kinds " cbf_cr " is defined below.
In the case of cbf_cr [x0] [y0] [trafoDepth] " being " 1 ", Cb transform blocks include 1 that rank is not 0 Conversion coefficient above.X0 and y0 represent prediction on the basis of the luma samples positioned at the upper left corner of image, positioned at object The coordinate of the luma samples in the upper left corner of block.TrafoDepth represents the current segmentation level of the CU in the block of transition coding Not.In the case where trafoDepth is " 0 ", block corresponds to CU.Be not present in cbf_cr [x0] [y0] [trafoDepth] and And in the case that PredMode is different from MODE_INTRA, cbf_cr [x0] [y0] [trafoDepth] value is considered as " 0 ".
That is, the position of Cb transform blocks is defined to include key element vertically and horizontally and on the basis of CU level With respect to the value of the three-dimensional arrangement including hierarchical depth (trafoDepth).
Figure 28 A and Figure 28 B are the tables for representing the result on cbf_luma, cbf_cb and cbf_cr.
Figure 28 A represent the result in the case of use (pattern 2) 1 adjacent block (only left piece condition L judgment value). Figure 28 B represent the result in the case of use (pattern 3) 0 adjacent block (upper block condition L and a left side are all not used).
In addition, the result and the 1st checking shown in Figure 28 A and Figure 28 B are same, the BD- of above-mentioned 4 test patterns is represented The increase and decrease of Rate values.In addition, the meaning of evaluation of estimate is same with the 1st checking.
According to the result, as a result with the skip_ of the checkings of split_coding_unit_flag and the 2nd of the above-mentioned 1st checking Flag is different, on cbf_luma, cbf_cb and the cbf_cr, as the pattern of context model in pattern 1 and pattern 2 or Do not occur BD-rate larger difference between pattern 3.
Thus, context control unit 142 and 242 is special in the environment of control parameter of multiple signal kinds is mixed It is unfavorable with above and below piece being determined as adjacent block not in the case where signal kinds are cbf_luma, cbf_cb and cbf_cr Text value.That is, context control unit 142 and 242 is cbf_luma, cbf_cb and cbf_cr feelings in the signal kinds of control parameter Under condition, use pattern 2 or pattern 3 determine context value.In other words, above-mentioned 1st species includes " split_coding_unit_ Flag " or " skip_flag ", above-mentioned 2nd species or the 3rd species include " cbf_luma ", " cbf_cb " or " cbf_cr ".By This, can be while the decline of BD-rate values be suppressed about the picture coding device and picture decoding apparatus of present embodiment Cut down memory usage amount.
In addition, in the case where comparing pattern 2 and pattern 3 for cbf_luma, cbf_cb and cbf_cr, they Do not occur larger difference in BD-rate.Therefore, it is more preferably use pattern 3 for cbf_luma, cbf_cb and cbf_cr. Thereby, it is possible to realize the reduction of further memory usage amount and the reduction for the treatment of capacity.
In addition, cbf_luma, cbf_cb and cbf_cr are the values of the three-dimensional arrangement with level.Therefore, such as in embodiment party In Fig. 9 B of formula 2 describe as, " adjoining " reference point according to the suitable of decoding order (including recurrence execution sequence, as follows) Sequence distance (distance) is different according to hierarchical depth.Accordingly, with respect to whether can using control parameter value or memory make Whether the reduction of consumption tells on, different according to the position in the hierarchical relationship of block.
Therefore, context control unit 142 and 242 can also change the decision standard of condition value according to hierarchical depth.For example, Context control unit 142 and 242 can also by the use of block of the condition value as level on the lower of the top block of (continuing to use) level bar Part value.
Also, context control unit 142 and 242 can also on this basis or individually consider the position with other sections Relation is put to change their standard.
In addition, can consider following specific deformation by above-mentioned checking.About whether the above-mentioned " dilution of generation Effect ", dependent on the study number of times for being applicable its condition.Generally for 4:2:The Y of brightness is represented in Y, U, the V such as 0 form, its sample This quantity is more than other two axles (U, V).Therefore, a border for making a distinction is (a) luma and (b) cv and cr group Border.
For example, it is also possible to pattern 3 is applicable to (a) cbf_luma, while being applicable (b) cbf_cb and cbf_cr other moulds Formula 2 or pattern 1.I.e. or, in the case where signal kinds are cbf_luma and signal kinds be cbf_cb or cbf_cr In the case of, context control unit 142 and 242 uses different conditional decision contexts.
Here, as long as the study of context control unit 142 and 242 number of times enough, increases (up and down more preferably when obtaining precision Text) condition.If in addition, the study number of times of context control unit 142 and 242 is few, preferably reducing context condition.Therefore, on Hereafter control unit 142 and 242 can also switch these conditions according to resolution.Also, context control unit 142 and 242 also may be used With according to form (4:2:Etc. 0) these conditions are switched.
In addition, cbf_luma, cbf_cr and cbf_cb whether there is the coefficient of brightness or aberration according to the depth representing of level. That is, cbf_luma, cbf_cr and cbf_cb, which correspond to, represents whether there is under the no_residual_data_flag of residual error data Position.Here, for the maximum TU that can be selected for CU size in CU≤PU≤TU relation, using no_ residual_data_flag.Specifically, no_residual_data_flag is used in TU uppermost level.Separately On the one hand, cbf_luma, cbf_cr and cbf_cb are used in than the no_residual_data_flag on the lower level of position.Such as There is no_residual_data_flag in fruit, then it represents that block henceforth does not have residual error data.Also, level is deeper, have The possibility of residual error data is bigger.Therefore, context control unit 142 and 242 can be by cbf_luma, cbf_cr and cbf_ Cb improves code efficiency using the information of level.In other words, context control unit 142 and 242 can also be according to process object The depth of the level of data unit belonging to the control parameter (cbf_luma, cbf_cr and cbf_cb) of block determines context.Separately On the one hand, no_residual_data_flag is the mark for being dependent on predictive mode, thus is preferably used not as context value Fixed value dependent on level.
Below, the result of above-mentioned all signal kinds is summarized expression.
Figure 29 A and Figure 29 B are to represent 4.1Intra, high-efficiency setting result (all signal kinds Class) block diagram.
Figure 30 A and Figure 30 B are to represent 4.3Random access, and high-efficiency setting result is (all Signal kinds) block diagram.
Figure 31 A and Figure 31 B are to represent 4.5Low delay, high-efficiency setting result (all signals Species) block diagram.
Figure 32 A and Figure 32 B are to represent 4.7Low delay, high-efficiency setting (only Pslices) post Shape figure.
Figure 33 A are the parameter sets for representing to distribute each control parameter any one pattern in above-mentioned 1~pattern of pattern 3 Table.In the example shown in Figure 33 A, to (the upper block of " split_coding_unit_flag " and " skip_flag " allocation model 1 And left piece all used), to " merge_flag ", " ref_idx ", " inter_pred_flag ", " mvd_l0 ", " mvd_l1 " And " mvd_lc ", " no_residual_data_flag ", " intra_chroma_pred_mode ", " cbf_luma ", " cbf_ Cb " and " cbf_cr " allocation models 3 (go up block and left piece is all not used).
Figure 33 B are the tables of the result when representing to use the parameter set shown in Figure 33 A.As shown in Figure 33 B, by using Parameter set shown in Figure 33 A, can cut down memory usage amount while the decline of code efficiency is suppressed.
Figure 34 A are the tables of one for representing another parameter set.In the example shown in Figure 34 A, to " split_ Coding_unit_flag " and " skip_flag " allocation models 1 (go up block and left piece is all used), to " intra_chroma_ Pred_mode ", " cbf_luma ", " cbf_cb " and " cbf_cr " allocation models 2 (being used only left piece), to " merge_ Flag ", " ref_idx ", " inter_pred_flag ", " mvd_l0 ", " mvd_l1 " and " mvd_lc ", " no_residual_ Data_flag " allocation models 3 (go up block and left piece are all not used).
Figure 34 B are the tables of the result when representing to use the parameter set shown in Figure 34 A.As illustrated in figure 34b, by using Parameter set shown in Figure 34 A, can cut down memory usage amount while the decline of code efficiency is suppressed.
More than, the picture coding device and picture decoding apparatus of embodiment for the present invention are illustrated, but The present invention is not limited to the embodiment.
For example, it is also possible to by about the picture coding device of above-mentioned embodiment 1~3, picture decoding apparatus and they At least a portion combination in the function of variation.
In addition, be entirely to be illustrated to specifically describe the present invention in the above-mentioned middle numeral used and logical value, this The numerical limit that invention is not illustrated.
In addition, the segmentation of the functional block in block diagram is one, it can also be realized multiple functional blocks as One function block, Or One function block is divided into it is multiple, or by a part function be transferred in other functional blocks.In addition it is also possible to be single Hardware or software by the function parallelization or time-division processing of multiple functional blocks with similar function.
In addition, holding in the method for encoding images of above-mentioned picture coding device and the picture decoding method of picture decoding apparatus The order of each step of row is illustrated or order other than the above to specifically describe the present invention.In addition, above-mentioned A part for step can also be with (parallel) execution simultaneously of other steps.
(embodiment 4)
By by for realizing dynamic image encoding method (method for encoding images) shown in the respective embodiments described above or fortune The program of the structure of moving picture decoding method (picture decoding method) recorded in storage medium, can be by the respective embodiments described above Shown processing is simply implemented in independent computer system.Storage medium is disk, CD, photomagneto disk, IC-card, partly led Body memory etc., as long as it is able to record that the medium of program just can be with.
And then, illustrate here the dynamic image encoding method (method for encoding images) that shows in the respective embodiments described above and The application examples of dynamic image decoding method (picture decoding method) and use its system.The system is characterized in that, with by The Image Coding that picture coding device using method for encoding images and the picture decoding apparatus using picture decoding method are constituted Decoding apparatus.On the other structures of system, can according to circumstances and suitably it change.
Figure 37 is the integrally-built figure for representing to realize the content provider system ex100 of content distribution service.By communication clothes The offer zoning of business is divided into desired size, be respectively arranged with each cell base station ex106 as stationary wireless stations, ex107、ex108、ex109、ex110。
Content provider system ex100 is on internet ex101 via ISP ex102 and telephone network Ex104 and base station ex107~ex110 are connected to computer ex111, PDA (PersonalDigital Assistant) Ex112, camera ex113, portable phone ex114, game machine ex115 etc. each equipment.
But, content provider system ex100 is not limited to structure as Figure 37, can also connect some factor combinations Connect.In addition it is also possible to which each equipment is connected directly between into telephone network not via base station ex107~ex110 as stationary wireless stations On ex104.In addition it is also possible to which each equipment is directly connected to each other via near radio etc..
Camera ex113 is the equipment for the moving image photography that can carry out digital camera etc., and camera ex116 is energy Enough carry out still image photographing, the equipment of moving image photography of digital camera etc..In addition, portable phone ex114 is GSM (Global System for Mobile Communications) mode, CDMA (Code Division Multiple Access) mode, W-CDMA (Wideband-Code Division Multiple Access) modes or LTE (Long Term Evolution) mode, HSPA (High Speed Packet Access) portable telephone or PHS (Personal Handyphone System) etc., it is which kind of can.
In content provider system ex100, by the way that camera ex113 etc. is connected via base station ex109, telephone network ex104 It is connected on streaming media server ex103, can relays live etc..In relaying at the scene, camera ex113 is used to user The content (such as the image at concert scene) of photography carries out coded treatment as described above in the embodiments (that is, being played a role as the picture coding device of the mode of the present invention), sends to streaming media server ex103.It is another Aspect, streaming media server ex103 is by the content-data sent to there is the client of request to carry out stream distribution.It is used as client , there are computer ex111, PDAex112, camera ex113, the portable electricity that can decode the data after above-mentioned coded treatment in end Talk about ex114, game machine ex115 etc..In each equipment of the data of distribution is received, by the data decoding process received again Existing (that is, being played a role as the picture decoding apparatus of the mode of the present invention).
In addition, the coded treatment of data of photography can both be carried out by camera ex113, can also by carry out data hair Send the streaming media server ex103 of processing to carry out, can also mutually share progress.Equally, the decoding process of the data of distribution had been both It can be carried out, can also be carried out by streaming media server ex103 by client, can also mutually share progress.In addition, not limiting In camera ex113, can also by by the camera ex116 rest images photographed and/or motion image data via computer Ex111 is sent to streaming media server ex103.Coded treatment in the case of this is by camera ex116, computer ex111, stream matchmaker Body server ex103 which carry out can, can also mutually share progress.
Handled in addition, the processing of these coding and decodings is general in the LSIex500 that computer ex111 or each equipment has. LSIex500 both can be single-chip or the structure being made up of multiple chips.Alternatively, it is also possible to by moving image encoding The software of decoding be encased in can by the readings such as computer ex111 some recording mediums (CD-ROM, floppy disk, hard disk etc.) In, use the software carry out coding and decoding processing.And then, can also in the case where portable phone ex114 is to carry camera The motion image data obtained by the camera is sent.Motion image data now is had by portable phone ex114 The data of LSIex500 coded treatments.
In addition it is also possible to be, streaming media server ex103 is multiple servers or multiple computers, is to disperse data Handle, record and distribute.
As above, in content provider system ex100, client can receive the data of coding and reproduce.So, In content provider system ex100, the information that client can send user is received, decodes, reproduced in real time, even not having The user for having special right or equipment can also realize personal broadcaster.
In addition, content provider system ex100 example is not limited to, as shown in figure 38, in digital broadcasting system Also at least dynamic image encoding device (picture coding device) or moving image solution of above-mentioned embodiment can be loaded in ex200 Some of code device (picture decoding apparatus).Specifically, in the ex201 of broadcasting station, music number will be multiplexed to image data Transmitted according to multiplex data obtained from grade via electric wave to communication or broadcasting satellite ex202.The image data is by above-mentioned each Data after the dynamic image encoding method coding illustrated in embodiment (that is, are compiled by the image of the mode of the present invention Data after code device code).The broadcasting satellite ex202 for receiving the data sends the electric wave of broadcast, can be to the electric wave The antenna ex204 for carrying out the family of satellite radio receiver receives the electric wave, passes through television set (receiver) ex300 or set top box (STB) multiplex data received is decoded and is rendered by ex217 etc. device (that is, is used as a mode of the invention Picture decoding apparatus plays a role).
In addition it is also possible to be that the multiplex data recorded in DVD, BD etc. recording medium ex215 is being read and solved Code encodes image data the reader/record for being multiplexed and writing in recording medium ex215 further according to situation and music signal Moving image decoding apparatus or dynamic image encoding device shown in the respective embodiments described above can be also installed in device ex218. In this case, can include the signal of video signal of reproduction on monitor ex219, there is the recording medium of multiplex data by record Ex215 can reproduce signal of video signal in other devices or system.In addition it is also possible to be to be connected to the line of cable television Moving image decoding apparatus is installed in set top box ex217 on the cable ex203 or antenna ex204 of satellite/terrestrial ripple broadcast, by it Shown with the monitor ex219 of television set.At this time it is also possible to not be to load moving image solution in television set in set top box Code device.
Figure 39 is to represent the dynamic image decoding method and moving image encoding using illustrating in the respective embodiments described above Television set (receiver) ex300 of method figure.Television set ex300 possesses via the antenna ex204 or line for receiving above-mentioned broadcast Cable ex203 etc. obtains or exported the tuner ex301 for the multiplex data that voice data has been multiplexed to image data, will receive The modulating/demodulating portion ex302 of coded data that is sent to outside of multiplex data demodulation or be modulated to and by the multiplexing after demodulation Data separating is image data, voice data or is multiplexed the image data encoded in signal processing part ex306, voice data Multiplexing/separation unit ex303.
In addition, television set ex300 possesses：With voice data, image data are decoded respectively or respective information is compiled The sound signal processing portion ex304 and signal of video signal processing unit ex305 of code (that is, are compiled as the image of the mode of the present invention Code device or picture decoding apparatus play a role) signal processing part ex306；With raising of exporting decoded voice signal The display part ex308 of the display of sound device ex307 and the decoded signal of video signal of display etc. output section ex309.And then, TV Machine ex300 possesses the interface portion ex317 with the operation inputting part ex312 of input for accepting user's operation etc..And then, television set Ex300 has control unit ex310, the power circuit part ex311 to each portion's supply electric power for merging each portion of control.Interface portion Ex317 can also be also connected in addition to operation inputting part ex312 with the external equipment with reader/writer ex218 etc. Bridge part ex313, the socket ex314 of recording medium ex216 for SD card etc. can be installed, for outer with hard disk etc. The driver ex315 of portion's recording medium connection, the modem ex316 being connected with telephone network etc..In addition, recording medium Ex216 is the knot for the record that can electrically enter row information by the semiconductor memory component of non-volatile/volatibility of harvesting Structure.Television set ex300 each portion is connected with each other via synchronous bus.
First, the knot that television set ex300 will be decoded, reproduced by the multiplex data obtained from outside such as antenna ex204 Structure is illustrated.Television set ex300 receives user's operation from remote control ex220 etc., based on the control unit with CPU etc. Ex310 control, the multiplex data demodulated by modulating/demodulating portion ex302 is separated with multiplexing/separation unit ex303.And then, electricity The voice data of separation is decoded with sound signal processing portion ex304 depending on machine ex300, by the image data signal of video signal of separation Processing unit ex305 is decoded using the coding/decoding method illustrated in the respective embodiments described above.Decoded voice signal, image are believed Exported number respectively from output section ex309 towards outside.Output when, can temporarily by these signals be stored into buffer ex318, In ex319 etc., so that voice signal and signal of video signal reproduced in synchronization.In addition, television set ex300 may not be from broadcast etc., And the multiplex data of coding is read from recording medium ex215, ex216 of magnetic optical disk, SD card etc..Then, to television set ex300 Illustrated by voice signal or signal of video signal coding, to the outside structure for sending or being written in recording medium etc..Television set Ex300 receives user's operation from remote control ex220 etc., based on control unit ex310 control, by sound signal processing portion Signal of video signal is used and illustrated in the respective embodiments described above by ex304 by sound signal encoding by signal of video signal processing unit ex305 Coding method coding.Voice signal after coding, signal of video signal are multiplexed with multiplexing/separation unit ex303, to outside output. In multiplexing, temporarily these signals can be stored into buffer ex320, ex321 etc., so that voice signal and signal of video signal Reproduced in synchronization.In addition, buffer ex318, ex319, ex320, ex321 can both possess as illustrated it is multiple or Share the structure of more than one buffer.And then, beyond diagram or, in such as modulating/demodulating portion ex302 or Also serve as between multiplexing/separation unit ex303 etc. avoiding the overflow of system, the buffer part of underflow and store data in a buffer.
In addition, television set ex300 from broadcast etc. or recording medium etc. in addition to obtaining voice data, image data, Can possess the structure for the AV inputs for accepting microphone or camera, coded treatment is carried out to the data obtained from them.Separately Outside, here, said television set ex300 as the structure that can carry out above-mentioned coded treatment, multiplexing and outside output It is bright but it is also possible to be, it is impossible to carry out these processing, and be only capable of carrying out above-mentioned reception, decoding process, the knot of outside output Structure.
In addition, by reader/writer ex218 from the case that multiplex data is read or write by recording medium, on State decoding process or coded treatment by television set ex300, reader/writer ex218 which carry out can or Television set ex300 and reader/writer ex218 mutually share progress.
As one, by information regeneration/record portion ex400's in the case of reading or write-in that data are carried out from CD In structure Biao Shi Figure 40.Information regeneration/record portion ex400 possess the unit ex401 of following explanation, ex402, ex403, ex404、ex405、ex406、ex407.Bare headed ex401 irradiates laser spot to the recording surface of the recording medium ex215 as CD And information is write, detect the reflected light of the recording surface from recording medium ex215 and read in information.Modulate record portion ex402 electricity The semiconductor laser being built in bare headed ex401 is driven gas, and the modulation of laser is carried out according to record data.Reproduce demodulation section Ex403 the reflected light from recording surface electrically will be detected as the photodetector being built in bare headed ex401 obtained from again Existing signal amplification, the signal component recorded in recording medium ex215 is separated and demodulated, the information required for reproducing.Buffering Device ex404 is by for recorded information in recording medium ex215 and temporarily kept from the recording medium ex215 information reproduced. Table motor ex405 rotates recording medium ex215.Servo control portion ex406 is while control table motor ex405 rotation driving one While making bare headed ex401 be moved to defined information track, the tracking process of laser spot is carried out.Systems control division ex407 is believed Cease the overall controls of reproduction/record portion ex400.The processing of above-mentioned read and write utilizes holding by systems control division ex407 Various information in buffer ex404, carry out further according to needs the generating of new information, it is additional and while make tune Record portion ex402 processed, reproduce demodulation section ex403, servo control portion ex406 coordinations, while being believed by bare headed ex401 The record of breath reproduces to realize.Systems control division ex407 is for example made up of microprocessor, by perform read write-in program come Perform their processing.
More than, it is assumed that bare headed ex401 is irradiated laser spot and is illustrated highly dense but it is also possible to be being carried out using near field light The structure of the record of degree.
The recording medium ex215 of CD schematic diagram is denoted as in Figure 41.In recording medium ex215 recording surface On, guide channel (ditch) is spirally formed with, in information track ex230, has been recorded beforehand through the change of the shape of ditch The address information of absolute position on indicating panel.The address information includes the record block for being used for being determined as the unit of record data The information of ex231 position, reads address and believes by reproducing information track ex230 in the device for being recorded and being reproduced Breath, can determine record block.In addition, recording medium ex215 includes data recording area ex233, inner region ex232, periphery Region ex234.The region used to record user data is data recording area ex233, is configured than data recorded area The inner region ex232 and outer region ex234 of domain ex233 inner circumferential sides or periphery are used for specific beyond the record of user data Purposes.The sound that information regeneration/record portion ex400 is encoded to such recording medium ex215 data recording area ex233 Sound data, image data or be multiplexed these data coded data read-write.
More than, it is illustrated exemplified by the CD for lifting 1 layer DVD, BD etc., but be not limited to these or many Layer construction, the CD being also able to record that beyond surface.In addition it is also possible to be to use different wave length in the identical place of disk The light recording information of color or layer from the different information of various angle recordings etc., carry out multidimensional record/reproduction construction CD.
, can also be by the car ex210 with antenna ex205 from satellite in addition, in digital broadcasting system ex200 Ex202 etc. receives data, the reproduction of moving image in the car ex210 automatic navigator ex211 having etc. display device.Separately Outside, automatic navigator ex211 structure can contemplate the structure that GPS receiving unit is added for example in the structure shown in Figure 39, It is also contemplated that same structure in computer ex111 and portable phone ex114 etc..
Figure 42 A are to represent the dynamic image decoding method and moving image encoding side using illustrating in the above-described embodiment The portable phone ex114 of method figure.Portable phone ex114 has by the antenna for the transmitting-receiving electric wave between the ex110 of base station Ex350, can filmed image, the camera section ex365 of rest image, display will photographed by camera section ex365 image, by The display part ex358 of the liquid crystal display of the decoded data such as the image that antenna ex350 is received etc..Portable phone ex114 The audio output unit ex357 of loudspeaker also with the main part comprising operated key portion ex366, for carrying out sound output etc., The sound input unit ex356 of microphone for carrying out sound input etc., preserve the image photographed, rest image, recording The memory portion of data or decoded data after the coding of sound or the image received, rest image, mail etc. The ex367 or socket ex364 as the interface portion between the same recording medium for preserving data.
And then, portable phone ex114 configuration example is illustrated using Figure 42 B.Portable phone ex114 is controlled for merging Fixture is for the display part ex358 and master control part ex360 in each portion of operated key portion ex366 main part, by power circuit part Ex361, operation input control unit ex362, signal of video signal processing unit ex355, camera interface portion ex363, LCD (Liquid Crystal Display：Liquid crystal display) control unit ex359, modulating/demodulating portion ex352, multiplexing/separation unit ex353, sound Signal processing part ex354, socket ex364, memory portion ex367 are connected with each other via bus ex370.
If power circuit part ex361 makes end of conversation and power key turn into opening by the operation of user, lead to Cross from battery pack to each portion's supply electric power, portable phone ex114 is activated in enabled state.
Controls of the portable phone ex114 based on the master control part ex360 with CPU, ROM and RAM etc., in voice call mould During formula, believe digital audio is transformed to by sound signal processing portion ex354 by the voice signal of sound input unit ex356 collection sounds Number, it is subjected to wave spectrum DIFFUSION TREATMENT with modulating/demodulating portion ex352, digitaltoanalogconversion is implemented by transmission/reception unit ex351 Through being sent by antenna ex350 after processing and frequency conversion process., will be by addition, portable phone ex114 is in voice calling mode The reception data that antenna ex350 is received are amplified and implement frequency conversion process and analog-to-digital conversion processing, use modulating/demodulating Portion ex352 carries out wave spectrum counter diffusion processing, is transformed to after analog sound data, is passed through by sound signal processing portion ex354 Exported by audio output unit ex357.
And then, in the case of sending Email in data communication mode, the operated key portion ex366 of main part will be passed through Deng operation input Email text data via operation input control unit ex362 to master control part ex360 send out.It is main Text data is carried out wave spectrum DIFFUSION TREATMENT by control unit ex360 with modulating/demodulating portion ex352, real by transmission/reception unit ex351 Apply after digitaltoanalogconversion processing and frequency conversion process, sent via antenna ex350 to base station ex110.Receiving electronics postal In the case of part, the substantially inversely processing of above-mentioned processing is performed to the data received, and be output to display part ex350.
In data communication mode, in the case of transmission image, rest image or image and sound, signal of video signal Processing unit ex355 compiles the signal of video signal supplied from camera section ex365 by the moving image shown in the respective embodiments described above Code method is compressed coding (that is, being played a role as the picture coding device of the mode of the present invention), after coding Image data is sent to multiplexing/separation unit ex353.In addition, sound signal processing portion ex354 by camera section ex365 to being clapped Encoded during photogra, rest image etc. with the voice signal of sound input unit ex356 collection sounds, by the sound after coding Sound data are sent to multiplexing/separation unit ex353.
Multiplexing/separation unit ex353 is by defined mode, to after the coding that is supplied from signal of video signal processing unit ex355 Voice data after image data and the coding supplied from sound signal processing portion ex354 is multiplexed, and its result is obtained Multiplex data carries out wave spectrum DIFFUSION TREATMENT with modulating/demodulating portion (modulation/demodulation circuit portion) ex352, by transmission/reception unit Ex351 is implemented after digitaltoanalogconversion processing and frequency conversion process, is sent via antenna ex350.
In the case of the data that the motion pictures files for being linked to homepage etc. are received in data communication mode, Huo Zhejie In the case of receiving the Email that addition of image or sound, in order to the multiplex data received via antenna ex350 Decoded, multiplexing/separation unit ex353 is divided into the bit stream and voice data of image data by the way that multiplex data is separated Bit stream, supplies the image data after coding to signal of video signal processing unit ex355 via synchronous bus ex370, and will coding Voice data afterwards is supplied to sound signal processing portion ex354.Signal of video signal processing unit ex355 by with the respective embodiments described above The corresponding dynamic image decoding method of shown dynamic image encoding method is decoded, and thus signal of video signal is decoded (that is, being played a role as the picture decoding apparatus of a mode of the present invention), via LCD control units ex359 from display part Ex358, which is shown, is for example linked to the image included in the motion pictures files of homepage, rest image.In addition, sound signal processing Portion ex354 is decoded to voice signal, and sound is exported from audio output unit ex357.
In addition, above-mentioned portable phone ex114 etc. terminal is same with television set ex300, except with encoder, decoder Beyond both transmitting-receiving type terminal, it is also contemplated that only the transmissions terminal of encoder, there was only 3 kinds of receiving terminal of decoder Installation form.In addition, in digital broadcasting system ex200, it is set to send, receives music data has been multiplexed in image data It is illustrated etc. obtained multiplex data, but has been multiplexed in addition to voice data the number of character data with image association etc. According to can also, be not multiplex data but image data in itself can also.
So, the dynamic image encoding method represented in the respective embodiments described above or dynamic image decoding method are used in In which kind of above-mentioned equipment, system can, by such manner, it is possible to the effect illustrated in the respective embodiments described above.
In addition, the present invention is not limited to such above-mentioned embodiment, the scope of the present invention can not departed from and carried out Various modifications or amendment.
(embodiment 5)
Can also be by by the dynamic image encoding method shown in the respective embodiments described above or device and foundation The dynamic image encoding method or device of MPEG-2, MPEG4-AVC, VC-1 etc. different standards as needed and are suitably cut Change, to generate image data.
Here, generation respectively according to multiple image datas of different standards in the case of, decoding when, it is necessary to select Corresponding to the coding/decoding method of each standard.But, due to the image data to be decoded can not be recognized according to which standard, so production The problem of life can not select appropriate coding/decoding method.
In order to solve the problem, the multiplex data that voice data etc. has been multiplexed in image data is used comprising expression image Structure of the data according to the identification information of which standard.Hereinafter, illustrate to include the fortune by showing in the respective embodiments described above The specific structure of multiplex data including moving picture encoding method or the image data of device generation.Multiplex data is MPEG- The digital stream of 2 transport stream forms.
Figure 43 is the figure for the structure for representing multiplex data.As shown in figure 43, multiplex data by by video flowing, audio stream, More than 1 in presentation graphic stream (PG), interactive graphic stream is multiplexed and is obtained.Video flowing represents the main image and pair of film Image, the secondary sound that audio stream (IG) represents the master voice part of film and mixed with the master voice, presentation graphic stream represents electricity The captions of shadow.Here, so-called main image, represents the common image being shown on picture, so-called subpictures are in main image The image shown with less picture.In addition, interactive graphic stream represents the dialogue by configuring GUI component on picture and making Picture.Video flowing is by the dynamic image encoding method or device that show in the respective embodiments described above, according to conventional The dynamic image encoding method or device code of the standards such as MPEG-2, MPEG4-AVC, VC-1.Audio stream by Dolby AC-3, Dolby Digital Plus, MLP, DTS, DTS-HD or linear PCM etc. mode are encoded.
Each stream included in multiplex data is identified by PID.For example, to the video flowing used in the image of film 0x1011 is distributed, to audio stream distribution 0x1100 to 0x111F, 0x1200 to 0x121F is distributed to Presentation Graphics, to interactive graphics (IG) Stream distribution 0x1400 to 0x141F, to video flowing distribution 0x1B00 to the 0x1B1F used in the subpictures of film, pair and lead Audio stream distribution 0x1A00 to the 0x1A1F used in the secondary sound of sound mix.
Figure 44 is to show schematically the figure how multiplex data is re-used.First, the video that will be made up of multiple frame of video Stream ex235, the audio stream ex238 being made up of multiple audio frames are transformed to PES packet sequence ex236 and ex239, and are transformed to respectively TS bags ex237 and ex240.Equally, presentation graphic stream ex241 and interactive graphics (IG) ex244 data are transformed to PES bag sequences respectively Ex242 and ex245 is arranged, then is transformed to TS bags ex243 and ex246.Multiplex data ex247 by these TS claddings by using 1 Constituted in stream.
Figure 45 represents how to preserve video flowing in PES packet sequences in more detail.Figure 45 the 1st segment table shows regarding for video flowing Frequency frame sequence.2nd segment table shows PES packet sequences.As shown in Figure 45 arrow yy1, yy2, yy3, yy4, multiple works in video flowing I pictures, B pictures, P pictures for Video Presentation Unit are divided by each picture and are saved in having for PES bags Imitate in load.Each PES bags have PES heads, in PES, preserve the PTS at the display moment as picture (Presentation Time-Stamp) and the DTS (Decoding Time-Stamp) for decoding the moment as picture.
Figure 46 represents the form for the TS bags being finally written in multiplex data.TS bags are by letters such as the PID with identification stream The bag for the 188 byte regular lengths that the TS payload of the TS heads of 4 bytes of breath and 184 bytes of preservation data is constituted, it is above-mentioned PES coatings are split and are saved in TS payload.In the case of bd-rom, the TP_ of 4 bytes is assigned for TS bags Extra_Header, constitutes the source bag of 192 bytes, is written in multiplex data.ATS has been recorded in TP_Extra_Header (Arrival_Time_Stamp) information such as.ATS represents transfer start time of the TS bags to the pid filter of decoder. In multiplex data, source bag is arranged as shown in Figure 46 hypomeres, and incremental number is referred to as SPN (source bags from the beginning of multiplex data Number).
In addition, in the TS bags that multiplex data is included, in addition to each stream of image, sound, captions etc., also PAT (Program Association Table)、PMT(Program Map Table)、PCR(Program Clock Reference) etc..Pat table shows that what the PMT used in multiplex data PID is, PAT itself PID is registered as 0. PMT has the PID and the attribute of stream corresponding with each PID of each stream of image, sound, the captions that multiplex data included etc. Information, also with the various descriptors on multiplex data.In descriptor, have and indicate copying for permission/license multiplex data Copy control information of shellfish etc..PCR is in order to obtain ATC (Arrival Time Clock) and the conduct of the time shaft as ATS The STC (System Time Clock) of PTS and DTS time shaft synchronization, possesses and transfers to decoder with PCR coatings The information of ATS corresponding STC times.
Figure 47 is the figure for the data configuration for explaining PMT.In PMT beginning, it is configured with and describes included in the PMT In the length of data etc. PMT heads.Face, is configured with multiple descriptors on multiplex data behind.Above-mentioned copy control Information etc. is recited as descriptor.After descriptor, the stream letter of multiple each streams on included in multiplex data is configured with Breath.Stream information is by attribute information (the frame speed of the stream type, the PID flowed, stream that record the voice compression codecs for recognizing stream Rate, aspect ratio etc.) flow descriptors constitute.There is the quantity flowed present in multiplex data in flow descriptors.
In the case where recorded in recording medium etc., above-mentioned multiplex data is remembered together with multiplex data information file Record.
Multiplex data information file is the management information of multiplex data as shown in figure 48, right one to one with multiplex data Should, it is made up of multiplex data information, stream attribute information and entry maps.
Multiplex data information as shown in figure 48, is made up of system velocity, reproduction start time, reproduction finish time.System The maximum of the pid filter to System Target Decoder described later of rate representation multiplex data transfers speed.Included in multiplexing The interval of ATS in data is set as turning into below system velocity.Reproduction start time is the frame of video of the beginning of multiplex data PTS, reproduction finish time is set as the value plus the reproduction interval of 1 frame amount to the PTS of the frame of video of the end of multiplex data.
Stream attribute information as shown in figure 49, the attribute on each stream included in multiplex data is registered with by each PID Information.Attribute information has by video flowing, audio stream, presentation graphic stream, interactive graphic stream and different information.Video stream attribute There is information the video flowing to be many by the resolution ratio of what kind of voice compression codecs compression, each image data for constituting video flowing Less, the information that aspect ratio is how many, frame rate is how many etc..There is audio stream attribute information the audio stream to be compiled by what kind of compression The letter that decoder compression, the channel number included in the audio stream be how many, is how many etc. corresponding to which kind of language, sample frequency Breath.These information are used in initialization of decoder before player reproduction etc..
In the present embodiment, using stream type in above-mentioned multiplex data, included in PMT.In addition, being situated between in record In the case that record has multiplex data in matter, the video flowing attribute information included in multiplex data information is used.Specifically, In the dynamic image encoding method or device shown in the respective embodiments described above, following steps or unit, the step or list are set Member is to the stream type or video flowing attribute information included in PMT, and setting expression is by showing in the respective embodiments described above Dynamic image encoding method or device generation image data intrinsic information.By the structure, it can be identified by upper State the dynamic image encoding method shown in each embodiment or the image data and image according to other standards of device generation Data.
In addition, the step of representing the dynamic image decoding method of present embodiment in Figure 50.In step exS100, from The stream type being included in PMT or the video flowing attribute information included in multiplex data information are obtained in multiplex data.Connect , in step exS101, judge that stream type or video flowing attribute information are indicated whether by the respective embodiments described above The multiplex data of dynamic image encoding method or the device generation shown.Also, it is being judged as stream type or video stream attribute letter In the case of breath is the multiplex data of dynamic image encoding method by showing in the respective embodiments described above or device generation, In step exS102, decoded by the dynamic image decoding method shown in the respective embodiments described above.In addition, in stream class Type or video flowing attribute information represent be standard according to conventional MPEG-2, MPEG4-AVC, VC-1 etc. multiplex data In the case of, in step exS103, decoded by the dynamic image decoding method according to conventional standard.
So, by setting new eigenvalue in stream type or video flowing attribute information, can interpolate that in decoding is It is no to be decoded by the dynamic image decoding method or device shown in the respective embodiments described above.Thus, it is being transfused to In the case of multiplex data according to different standards, appropriate coding/decoding method or device can be also selected, so can not send out Life is mistakenly decoded.In addition, by the dynamic image encoding method shown in the present embodiment or device or motion diagram As coding/decoding method or device are used in any of the above described equipment, system.
(embodiment 6)
Dynamic image encoding method and device, dynamic image decoding method and the device shown in the respective embodiments described above Typically can be by the LSI realizations as integrated circuit.As one, the LSIex500 of 1 chip knot is represented in Figure 51 Structure.LSIex500 possess the unit ex501 of following explanation, ex502, ex503, ex504, ex505, ex506, ex507, ex508, Ex509, each unit is via bus ex510 connections.Power circuit part ex505 by being opening in power supply in the case of to each Portion's supply electric power, is activated in enabled state.
For example in the case where carrying out coded treatment, LSIex500 is based on having CPUex502, Memory Controller Ex503, stream controller ex504, driving frequency control unit ex512 etc. control unit ex501 control, pass through AV I/Oex509 AV signals are inputted from microphone ex117 and camera ex113 etc..The AV signals being transfused to are temporarily stored within SDRAM etc. outside Memory ex511 in.Based on control unit ex501 control, by the data of storage according to treating capacity and processing speed suitably Be divided into it is how inferior, to signal processing part ex507 send, in signal processing part ex507 carry out voice signal coding and/or shadow As the coding of signal.Here, the coded treatment of signal of video signal is the coded treatment illustrated in the respective embodiments described above.In signal In processing unit ex507, the processing of the image data multiplexing of the voice data for being about to coding and coding etc. is entered always according to situation, from I/Oex506 is flowed to export to outside.The bit of the output is flowed into base station ex107 transmissions or is written in recording medium ex215. In addition, multiplexing when, can temporarily by data storage into buffer ex508 so that its synchronization.
In addition, in above-mentioned, if memory ex511 is illustrated for the structure of LSIex500 outside, but it is also possible to It is included in the structure in LSIex500 inside.Buffer ex508 is also not limited to one, can also possess multiple bufferings Device.In addition, LSIex500 can both form 1 chip, multiple chips can also be formed.
In addition, in above-mentioned, it is assumed that control unit ex510 has CPUex502, Memory Controller ex503, stream controller Ex504, driving frequency control unit ex512 etc., but control unit ex510 structure is not limited to the structure.For example, it is also possible to be Signal processing part ex507 is also equipped with CPU structure.By being also provided with CPU, Neng Goujin in signal processing part ex507 inside One step improves processing speed.In addition, possessing signal processing part ex507 or as letter as other examples or CPUex502 The structure in such as sound signal processing portion of a number processing unit ex507 part.In this case, control unit ex501 is Possesses the structure with signal processing part ex507 or part thereof of CPUex502.
In addition, being set to LSI here, but according to the difference of integrated level, also there are referred to as IC, system LSI, super (super) LSI, superfine (ultra) LSI situation.
In addition, the method for integrated circuit is not limited to LSI, it can also be realized by special circuit or general processor. The FPGA (Field Programmable Gate Array) that can be programmed after LSI manufactures can also be utilized or can be reconstructed The reconfigurable processor of connection and the setting of circuit unit inside LSI.
And then, if occurring the integrated circuit instead of LSI because of the other technologies of the progress of semiconductor technology or derivation Technology, then can certainly use the technology carry out functional module it is integrated.It is likely to be application of biotechnology etc..
(embodiment 7)
In the image data for generating the dynamic image encoding method by showing in the respective embodiments described above or device In the case of decoding, it is contemplated that with by the feelings according to the image data of the standards such as conventional MPEG-2, MPEG4-AVC, VC-1 Condition can increase compared to treating capacity.Therefore, it is necessary to be set as than by the image data solution according to conventional standard in LSIex500 The higher driving frequency of driving frequency of CPUex502 during code.But, if setting driving frequency high, occur consumption electricity The problem of power is uprised.
In order to solve the problem, television set ex300, LSIex500 etc. moving image decoding apparatus are using identification image number According to according to which standard and according to the structure of standard switching driving frequency.Figure 52 represents the structure ex800 of present embodiment.Drive Dynamic frequency switching part ex803 is in the dynamic image encoding method or dress that image data is by showing in the respective embodiments described above In the case of putting generation, driving frequency is set to height.Also, to the moving image that execution is shown in the respective embodiments described above The decoding process portion ex801 of coding/decoding method indicates to decode image data.On the other hand, it is according to conventional mark in image data It is the dynamic image encoding method by showing in the respective embodiments described above with image data in the case of accurate image data Or the situation of the data of device generation is compared, and driving frequency is set to low.Also, to the decoding process according to conventional standard Portion ex802 indicates to decode image data.
More particularly, driving frequency switching part ex803 by Figure 51 CPUex502 and driving frequency control unit ex512 structures Into.In addition, performing decoding process portion ex801, Yi Jiyi of the dynamic image decoding method shown in the respective embodiments described above The decoding process portion ex802 of standard past according to this corresponds to Figure 51 signal processing part ex507.CPUex502 recognizes image data According to which standard.Also, based on the signal from CPUex502, driving frequency control unit ex512 setting driving frequencies.This Outside, based on the signal from CPUex502, signal processing part ex507 carries out the decoding of image data.Here it is possible to consider in shadow As using the identification information described in for example in embodiment 5 in the identification of data.On identification information, it is not limited in reality The information described in mode 5 is applied, as long as it can recognize that image data just can be with according to the information of which standard.For example, in base The external signal that television set is still used in disk etc. is used in identification image data, image data can be recognized according to which In the case of standard, it can also be identified based on such external signal.In addition, the selection example of CPUex502 driving frequency Such as it is contemplated that the standard by image data as shown in figure 50 is set up corresponding look-up table with driving frequency and carried out.By look-up table In the internal storage for being pre-reserved to buffer ex508 or LSI, CPUex502 can select to drive by referring to the look-up table Dynamic frequency.
The step of Figure 53 represents to implement the method for present embodiment.First, in step exS200, in signal processing part In ex507, identification information is obtained from multiplex data.Then, in step exS201, in CPUex502, based on identification letter Breath identification image data whether be coding method by showing in the respective embodiments described above or device generation data.In shadow In the case of being the data generated by the coding method shown in the respective embodiments described above or device as data, in step In exS202, CPUex502 sends to driving frequency control unit ex512 and driving frequency is set to high signal.Also, driving It is set as high driving frequency in dynamic frequency control unit ex512.On the other hand, represent be according to conventional MPEG-2, In the case of the image data of MPEG4-AVC, VC-1 etc. standard, in step exS203, CPUex502 is to driving frequency Control unit ex512 sends and driving frequency is set to low signal.Also, in driving frequency control unit ex512, be set as with Image data is the situation of the data of coding method by showing in the respective embodiments described above or device generation compared to lower Driving frequency.
And then, interlocked and changed to LSIex500 or the application of the device including LSIex500 by the switching with driving frequency Voltage, thus, it is possible to further improve power savings.For example, driving frequency is set to it is low in the case of, therewith, can be with Consider compared with driving frequency is set into high situation, by LSIex500 or the device including LSIex500 apply voltage It is set to low.
In addition, driving frequency is set in the case that as long as treating capacity of the establishing method of driving frequency in decoding is big High, in decoding treating capacity it is small in the case of by driving frequency be set to it is low just can be to be not limited to above-mentioned setting Method.For example, it may be considered that being more than in the treating capacity that will be decoded according to the image data of MPEG4-AVC standards will be by upper In the case of the treating capacity for the image data decoding for stating the dynamic image encoding method shown in each embodiment or device generation, It is driven the setting of frequency on the contrary with above-mentioned situation.
And then, the establishing method of driving frequency is not limited to the structure for making driving frequency low.For example, it is also possible to consider Identification information is the image data that dynamic image encoding method or device by showing in the respective embodiments described above are generated In the case of, by LSIex500 or the device including LSIex500 apply voltage be set to height, represent be according in the past , will be to LSIex500 or including LSIex500 in the case of the image data of MPEG-2, MPEG4-AVC, VC-1 etc. standard Device apply voltage be set to it is low.In addition, as another example, it is also contemplated that representing it is by above-mentioned in identification information In the case of the image data of dynamic image encoding method or the device generation shown in each embodiment, do not make CPUex502's Driving stops, and is the situation of the image data of the standard according to conventional MPEG-2, MPEG4-AVC, VC-1 etc. in expression Under, due to having more than needed in processes, so suspending CPUex502 driving.It is also contemplated that representing it is to pass through in identification information In the case of the image data of dynamic image encoding method or the device generation shown in the respective embodiments described above, as long as also existing Having to have more than needed in processing then suspends CPUex502 driving.In this case, it is possible to consider with representing it is according in the past The situation of the image data of MPEG-2, MPEG4-AVC, VC-1 etc. standard is compared, and dwell time is set to short.
So, according to the standard of image data institute foundation come switching driving frequency, thus, it is possible to realize energy-saving technology.In addition, In the case where driving LSIex500 or device including LSIex500 using battery, battery can be extended with economize on electricity Life-span.
(embodiment 8)
In the above-mentioned equipment such as television set, portable phone, system, multiple shadows according to different standards are transfused to sometimes As data.So, also it can decode that in order that obtaining in the case where being transfused to multiple image datas according to different standards, LSIex500 signal processing part ex507 needs to correspond to multiple standards.But, if be used alone corresponding to each standard Signal processing part ex507, then occur LSIex500 circuit scale become it is big, in addition cost increase the problem of.
In order to solve the problem, using by for performing the dynamic image decoding method shown in the respective embodiments described above Decoding process portion and according to conventional MPEG-2, MPEG4-AVC, VC-1 etc. standard decoding process portion a part altogether Structure.Figure 55 A ex900 represents the configuration example.For example, the moving picture decoding side shown in the respective embodiments described above Method and according to MPEG4-AVC standards dynamic image decoding method in entropy code, re-quantization, deblocking filter, motion compensation etc. Processing in some process content it is common.It is contemplated that following structure：On common process content, share and correspond to The decoding process portion ex902 of MPEG4-AVC standards, the of the invention mode on not corresponding to MPEG4-AVC standards Specific other process contents, use special decoding process portion ex901.Particularly, the present invention is in terms of motion compensation With feature, it can be considered to for example using special decoding process portion ex901 for motion compensation, in addition Some in entropy decoding, re-quantization or whole processing, share decoding process portion.On sharing for decoding process portion, It can be following structure：On common process content, the motion diagram for being used for performing and showing in the respective embodiments described above is shared As the decoding process portion of coding/decoding method, on process content specific to MPEG4-AVC standards, special decoding process is used Portion.
In addition, representing another example shared by a part is handled with Figure 55 B ex1000.In this embodiment, using using with The corresponding special decoding process portion ex1001 of process content specific to the mode of the present invention and with other marks in the past The accurate specific corresponding special decoding process portion ex1002 of process content and the motion diagram with a mode in the present invention As coding/decoding method shared decoding corresponding with process content common in other dynamic image decoding methods of standard in the past Processing unit ex1003 structure.Here, special decoding process portion ex1001, ex1002 are not necessarily of the invention one Mode or other process content specific to standard in the past and become privileged, can be able to carry out other general places The structure of reason.It is further possible to install the structure of present embodiment by LSIex500.
So, for the dynamic image decoding method in a mode of the invention and the moving image solution of conventional standard Common process content, shares decoding process portion, thus, it is possible to reduce LSI circuit scale and reduce cost in code method.
Industrial applicability
The present invention can be applied in method for encoding images, picture decoding method, picture coding device and picture decoding apparatus In, be particular enable to carry out the method for encoding images of arithmetic coding and arithmetic decoding, picture decoding method, picture coding device and Utilized in picture decoding apparatus.
Label declaration
The control unit of 100 picture coding device 101
102 difference portions 103 are converted and quantization unit
The re-quantization of 104 variable-length encoding part 105 and inverse transformation portion
106th, 206 adders 107,207 face interior prediction portions
108th, prediction section 109,209 is switched between 208 faces
The residual signals of 121 received image signal 122,125,225
123rd, the bit stream of 223 quantization transform coefficient 124
The prediction image signal of 126 restored image signal 127,128,129,227,228
130th, the binaryzation portion of 230 control parameter 141
142nd, the two-value arithmetic coding portion of 242 context control unit 143
151st, 251 two-values arrange 152,252 context indexs
The control unit of 200 picture decoding apparatus 201
Re-quantization portion of 202 variable-length decoding portion 204
The orthogonal transform coefficient of 205 inverse transformation portion 224
The picture signal of 226 decoded image signal 229
241 two-value arithmetic decoding portions of intermediate portion 243

Claims (7)

1. a kind of method for encoding images, uses arithmetic coding, it is characterised in that
Including：
Binarization step, passes through the control parameter binaryzation by block is dealt with objects, generation two-value arrangement；
Context rate-determining steps, determine the context used above-mentioned process object block in multiple contexts；And
Arithmetic coding step, arithmetic coding, generation ratio are carried out to the arrangement of above-mentioned two-value by using the above-mentioned context determined Special sequence；
In above-mentioned context rate-determining steps,
Judge the signal kinds of the control parameter of above-mentioned process object block；
In the case where above-mentioned signal kinds are the 1st species, using using with above-mentioned process object block abut left piece and upper piece Encoded control parameter the 1st condition, determine above-mentioned context；
In the case where above-mentioned signal kinds are with the above-mentioned 1st diverse 2nd species, using it is unfavorable with it is above-mentioned upper piece and on State left piece of encoded control parameter and using the 2nd condition of fixed value set in advance, determine above-mentioned context；
Above-mentioned 1st species is to represent whether above-mentioned process object block is divided into multiple pieces of mark or indicates whether above-mentioned place The mark that reason object block is skipped；
Above-mentioned 2nd species is the motion vector composition and the mark and table of the difference of its predicted value for representing above-mentioned process object block Show whether above-mentioned process object block includes the mark of the coefficient of brightness and aberration.
2. method for encoding images as claimed in claim 1, it is characterised in that
In above-mentioned context rate-determining steps, the also position based on above-mentioned process object block, whether judge being capable of profit in coding With above-mentioned upper piece control parameter；
The above-mentioned context of above-mentioned 2nd conditional decision can not used in the case of control parameter using block on above-mentioned.
3. method for encoding images as claimed in claim 2, it is characterised in that
In above-mentioned context rate-determining steps, in the case where above-mentioned process object block belongs to slice boundaries, it is judged as in coding Shi Buneng utilizes above-mentioned upper piece control parameter.
4. method for encoding images as claimed in claim 2, it is characterised in that
In above-mentioned context rate-determining steps, the level of the data unit according to belonging to the control parameter of above-mentioned process object block Depth judges whether the control parameter of block on above-mentioned can be utilized in coding.
5. method for encoding images as claimed in claim 1, it is characterised in that
In above-mentioned context rate-determining steps, the value of the control parameter also based on the 1st unit, contrast above-mentioned 1st unit it is small the 2nd The control parameter of unit, switching is using the above-mentioned context of above-mentioned 1st conditional decision or using in above-mentioned 2nd conditional decision Hereafter.
6. a kind of picture coding device, uses arithmetic coding, it is characterised in that
Including：
Binaryzation portion, passes through the control parameter binaryzation by block is dealt with objects, generation two-value arrangement；
Context control unit, determines the context used the arithmetic coding for dealing with objects block in multiple contexts；And
Arithmetic coding portion, carries out arithmetic coding to the arrangement of above-mentioned two-value by using the above-mentioned context determined, generates bit Sequence；
Above-mentioned context control unit,
Judge the signal kinds of the control parameter of above-mentioned process object block；
In the case where above-mentioned signal kinds are the 1st species, using using with above-mentioned process object block abut left piece and upper piece Encoded control parameter the 1st condition, determine above-mentioned context；
In the case where above-mentioned signal kinds are with the above-mentioned 1st diverse 2nd species, using it is unfavorable with it is above-mentioned upper piece and on State left piece of encoded control parameter and using the 2nd condition of fixed value set in advance, determine above-mentioned context；
Above-mentioned 1st species is to represent whether above-mentioned process object block is divided into multiple pieces of mark or indicates whether above-mentioned place The mark that reason object block is skipped；
Above-mentioned 2nd species is the motion vector composition and the mark and table of the difference of its predicted value for representing above-mentioned process object block Show whether above-mentioned process object block includes the mark of the coefficient of brightness and aberration.
7. a kind of image encoding/decoding device, it is characterised in that including：
Use the picture coding device of arithmetic coding；And
Use the picture decoding apparatus of arithmetic decoding；
Above-mentioned picture coding device includes：
Binaryzation portion, by the way that by the 1st control parameter binaryzation of the 1st process object block, the 1st two-value of generation is arranged；
1st context control unit, determined in multiple 1st contexts to the 1st process object block arithmetic coding use the 1st on Hereafter；And
Arithmetic coding portion, arithmetic coding is carried out by using above-mentioned 1st context determined to the arrangement of above-mentioned 1st two-value, raw Into the 1st bit sequence；
Above-mentioned 1st context control unit,
Judge the 1st signal kinds of the 1st control parameter of above-mentioned 1st process object block；
In the case where above-mentioned 1st signal kinds are the 1st species, the 1st left side that block is abutted is dealt with objects using using with the above-mentioned 1st 1st condition of the 1st encoded control parameter of block on block and the 1st, determines above-mentioned 1st context；
In the case where above-mentioned 1st signal kinds are with the above-mentioned 1st diverse 2nd species, using unfavorable with the above-mentioned 1st In 2nd condition of block and above-mentioned 1st left piece of the 1st encoded control parameter and utilization the 1st fixed value set in advance, decision State the 1st context；
Above-mentioned 1st species is to represent whether above-mentioned 1st process object block is divided into multiple pieces of mark or indicate whether will be upper State the mark that the 1st process object block is skipped；
Above-mentioned 2nd species be the difference of the motion vector composition and its predicted value that represent above-mentioned 1st process object block mark and Represent whether above-mentioned 1st process object block includes the mark of the coefficient of brightness and aberration；
Above-mentioned picture decoding apparatus includes：
2nd context control unit, determined in multiple 2nd contexts to the 2nd process object block arithmetic decoding use the 2nd on Hereafter；
Arithmetic decoding portion, will be by corresponding with the above-mentioned 2nd process object block obtained from the 2nd control parameter progress arithmetic coding The 2nd bit sequence, use above-mentioned 2nd context that is determined to carry out arithmetic decoding, thus the 2nd two-value arranged and restored；With And
Intermediate portion, by the way that above-mentioned 2nd two-value is arranged into intermediate, above-mentioned 2nd control parameter is restored；
Above-mentioned 2nd context control unit,
Judge the 2nd signal kinds of the 2nd control parameter of above-mentioned 2nd process object block；
In the case where above-mentioned 2nd signal kinds are the 3rd species, the 2nd left side that block is abutted is dealt with objects using using with the above-mentioned 2nd 3rd condition of decoded 2nd control parameter of block on block and the 2nd, determines above-mentioned 2nd context；
In the case where above-mentioned 2nd signal kinds are with the above-mentioned 3rd diverse 4th species, using unfavorable with the above-mentioned 2nd In 4th condition of block and above-mentioned 2nd left piece of decoded 2nd control parameter and utilization the 2nd fixed value set in advance, decision State the 2nd context；
Above-mentioned 3rd species is to represent whether above-mentioned 2nd process object block is divided into multiple pieces of mark or indicate whether will be upper State the mark that the 2nd process object block is skipped；
Above-mentioned 4th species be the difference of the motion vector composition and its predicted value that represent above-mentioned 2nd process object block mark and Represent whether above-mentioned 2nd process object block includes the mark of the coefficient of brightness and aberration.
CN201710263791.0A 2011-06-30 2012-06-22 Image encoding method, image encoding device, and image encoding/decoding device Active CN107181948B (en)

Applications Claiming Priority (3)

Application Number Priority Date Filing Date Title
US201161502992P 2011-06-30 2011-06-30
US61/502,992 2011-06-30
CN201280025858.2A CN103583048B (en) 2011-06-30 2012-06-22 Image decoding method, image encoding method, image decoding device, image encoding device, and image encoding/decoding device

Related Parent Applications (1)

Application Number Title Priority Date Filing Date
CN201280025858.2A Division CN103583048B (en) 2011-06-30 2012-06-22 Image decoding method, image encoding method, image decoding device, image encoding device, and image encoding/decoding device

Publications (2)

Publication Number Publication Date
CN107181948A true CN107181948A (en) 2017-09-19
CN107181948B CN107181948B (en) 2020-03-13

Family

ID=47390772

Family Applications (2)

Application Number Title Priority Date Filing Date
CN201280025858.2A Active CN103583048B (en) 2011-06-30 2012-06-22 Image decoding method, image encoding method, image decoding device, image encoding device, and image encoding/decoding device
CN201710263791.0A Active CN107181948B (en) 2011-06-30 2012-06-22 Image encoding method, image encoding device, and image encoding/decoding device

Family Applications Before (1)

Application Number Title Priority Date Filing Date
CN201280025858.2A Active CN103583048B (en) 2011-06-30 2012-06-22 Image decoding method, image encoding method, image decoding device, image encoding device, and image encoding/decoding device

Country Status (15)

Country Link
US (10) US8792739B2 (en)
EP (2) EP3944617A1 (en)
JP (5) JP5905884B2 (en)
KR (1) KR102007047B1 (en)
CN (2) CN103583048B (en)
AU (1) AU2012277220B2 (en)
BR (1) BR112013030562B1 (en)
CA (1) CA2837537C (en)
ES (1) ES2908152T3 (en)
MX (2) MX339141B (en)
MY (1) MY167090A (en)
PL (1) PL2728869T3 (en)
RU (1) RU2597473C2 (en)
TW (1) TWI564819B (en)
WO (1) WO2013001770A1 (en)

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113170122A (en) * 2018-12-01 2021-07-23 北京字节跳动网络技术有限公司 Parameter derivation for intra prediction
US11595687B2 (en) 2018-12-07 2023-02-28 Beijing Bytedance Network Technology Co., Ltd. Context-based intra prediction
US11729405B2 (en) 2019-02-24 2023-08-15 Beijing Bytedance Network Technology Co., Ltd. Parameter derivation for intra prediction
US11930185B2 (en) 2018-11-06 2024-03-12 Beijing Bytedance Network Technology Co., Ltd. Multi-parameters based intra prediction

Families Citing this family (37)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP5914962B2 (en) * 2010-04-09 2016-05-11 ソニー株式会社 Image processing apparatus and method, program, and recording medium
EP3907999B1 (en) 2010-09-02 2023-11-22 LG Electronics, Inc. Inter prediction
USRE47366E1 (en) 2011-06-23 2019-04-23 Sun Patent Trust Image decoding method and apparatus based on a signal type of the control parameter of the current block
CN103609113B (en) 2011-06-23 2017-03-08 太阳专利托管公司 Picture decoding method and picture decoding apparatus
CA2830046C (en) 2011-06-24 2018-09-04 Panasonic Corporation Image decoding method, image coding method, image decoding apparatus, image coding apparatus, and image coding and decoding apparatus
EP2725792B1 (en) * 2011-06-24 2018-08-22 Sun Patent Trust Image decoding method and image decoding device
CA2842646C (en) 2011-06-27 2018-09-04 Panasonic Corporation Image decoding method, image coding method, image decoding apparatus, image coding apparatus, and image coding and decoding apparatus
RU2602671C2 (en) 2011-06-28 2016-11-20 Сан Пэтент Траст Image decoding method, image encoding method, image decoding device, image encoding device, and image encoding and decoding device
KR102006032B1 (en) 2011-06-29 2019-07-31 선 페이턴트 트러스트 Image decoding method, image encoding method, image decoding device, image encoding device, and image encoding/decoding device
AU2012277220B2 (en) 2011-06-30 2016-08-11 Sun Patent Trust Image decoding method, image coding method, image decoding apparatus, image coding apparatus, and image coding and decoding apparatus
WO2013001769A1 (en) 2011-06-30 2013-01-03 パナソニック株式会社 Image decoding method, image encoding method, image decoding device, image encoding device, and image encoding/decoding device
KR20140034209A (en) 2011-07-11 2014-03-19 파나소닉 주식회사 Image decoding method, image encoding method, image decoding apparatus, image encoding apparatus, and image encoding/decoding apparatus
US9948938B2 (en) * 2011-07-21 2018-04-17 Texas Instruments Incorporated Methods and systems for chroma residual data prediction
US9571833B2 (en) * 2011-11-04 2017-02-14 Nokia Technologies Oy Method for coding and an apparatus
JP5485969B2 (en) 2011-11-07 2014-05-07 株式会社Ｎｔｔドコモ Moving picture predictive coding apparatus, moving picture predictive coding method, moving picture predictive coding program, moving picture predictive decoding apparatus, moving picture predictive decoding method, and moving picture predictive decoding program
US9185405B2 (en) 2012-03-23 2015-11-10 Qualcomm Incorporated Coded block flag inference in video coding
US9473769B2 (en) * 2012-06-08 2016-10-18 Texas Instruments Incorporated Method and system for reducing slice header parsing overhead in video coding
US9936200B2 (en) 2013-04-12 2018-04-03 Qualcomm Incorporated Rice parameter update for coefficient level coding in video coding process
JP6118199B2 (en) * 2013-07-09 2017-04-19 シャープ株式会社 Image decoding apparatus, image encoding apparatus, image decoding method, image encoding method, and computer-readable recording medium
US10021419B2 (en) * 2013-07-12 2018-07-10 Qualcomm Incorported Rice parameter initialization for coefficient level coding in video coding process
US9294766B2 (en) 2013-09-09 2016-03-22 Apple Inc. Chroma quantization in video coding
JP2015185897A (en) * 2014-03-20 2015-10-22 パナソニックＩｐマネジメント株式会社 Image encoding method and device
KR20170128390A (en) * 2015-03-19 2017-11-22 엘지전자 주식회사 Method for processing video signals and apparatus therefor
WO2016183814A1 (en) * 2015-05-20 2016-11-24 Mediatek Singapore Pte. Ltd. Coded block flag coding using cross-component correlation
EP3125108A1 (en) * 2015-07-31 2017-02-01 ARM Limited Vector processing using loops of dynamic vector length
JP6565885B2 (en) * 2016-12-06 2019-08-28 株式会社Ｊｖｃケンウッド Image encoding device, image encoding method, image encoding program, image decoding device, image decoding method, and image decoding program
KR102601268B1 (en) * 2017-02-24 2023-11-10 주식회사 케이티 Method and apparatus for processing a video signal
US10609414B2 (en) * 2017-05-08 2020-03-31 Qualcomm Incorporated Context modeling for transform coefficient coding
JP7168591B2 (en) * 2018-01-26 2022-11-09 パナソニック インテレクチュアル プロパティ コーポレーション オブ アメリカ Three-dimensional data encoding method, three-dimensional data decoding method, three-dimensional data encoding device, and three-dimensional data decoding device
CN111788829B (en) * 2018-02-28 2022-02-08 三星电子株式会社 Encoding method and device, and decoding method and device
WO2019216325A1 (en) 2018-05-09 2019-11-14 Sharp Kabushiki Kaisha Systems and methods for performing motion vector prediction using a derived set of motion vectors
AU2018233042A1 (en) 2018-09-21 2020-04-09 Canon Kabushiki Kaisha Method, apparatus and system for encoding and decoding a tree of blocks of video samples
JP2022504753A (en) * 2018-10-10 2022-01-13 インターデイジタル ヴィーシー ホールディングス インコーポレイテッド Affine mode signaling in video encoding and decoding
WO2020143643A1 (en) * 2019-01-07 2020-07-16 Beijing Bytedance Network Technology Co., Ltd. Control method for merge with mvd
WO2020171681A1 (en) * 2019-02-19 2020-08-27 주식회사 윌러스표준기술연구소 Intra prediction-based video signal processing method and device
WO2020256511A1 (en) * 2019-06-20 2020-12-24 엘지전자 주식회사 Video or image coding based on mapping of luma samples and scaling of chroma samples
BR112022008385A2 (en) * 2019-11-01 2022-07-12 Lg Electronics Inc DEVICE AND IMAGE ENCODING METHOD BASED ON TRANSFORMED TO THE SAME

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP1827030A2 (en) * 2006-01-10 2007-08-29 Samsung Electronics Co., Ltd. Method and apparatus for changing codec to reproduce video and/or audio data streams encoded by different codecs within a channel
CN101600104A (en) * 2008-06-02 2009-12-09 联发科技股份有限公司 The decoding unit of context adaptive binary arithmetic coding and coding/decoding method
CN101815224A (en) * 2005-07-22 2010-08-25 三菱电机株式会社 Picture coding device and method and picture decoding apparatus and method
CN101873492A (en) * 2002-08-08 2010-10-27 松下电器产业株式会社 Image encoding method and device

Family Cites Families (135)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPS53142227A (en) 1977-05-18 1978-12-11 Toshiba Shiyashin Youhin Kk Light control type electrooflashing device
JPS60125210A (en) 1983-12-08 1985-07-04 Agency Of Ind Science & Technol Improvement in preparation of polyimide membrane
JPS6225405A (en) 1985-07-26 1987-02-03 株式会社日立製作所 Temperature detection mechanism for resistor
JPS6225405U (en) 1985-07-30 1987-02-16
JPS638448A (en) 1986-06-27 1988-01-14 Nippon Synthetic Chem Ind Co Ltd:The Resin composition
JPS638448U (en) 1986-07-03 1988-01-20
TW330976B (en) 1995-12-18 1998-05-01 Lg Electronics Inc Constant temperature dehumidifying device for air regulator and control method thereof
EP1404133A4 (en) 2001-06-29 2010-04-21 Ntt Docomo Inc Image encoder, image decoder, image encoding method, and image decoding method
US7457359B2 (en) 2001-09-26 2008-11-25 Mabey Danny L Systems, devices and methods for securely distributing highly-compressed multimedia content
US7457358B2 (en) 2001-09-26 2008-11-25 Interact Devices, Inc. Polymorphic codec system and method
BR0206308B1 (en) 2001-11-06 2018-01-09 Matsushita Electric Ind Co Ltd "encoding method for encoding an image"
CN101448162B (en) 2001-12-17 2013-01-02 微软公司 Method for processing video image
JP2004088722A (en) 2002-03-04 2004-03-18 Matsushita Electric Ind Co Ltd Motion picture encoding method and motion picture decoding method
KR20040099100A (en) 2002-04-19 2004-11-26 마쯔시다덴기산교 가부시키가이샤 Method of calculating motion vector
EP3525464B1 (en) 2002-04-19 2021-03-03 Panasonic Intellectual Property Corporation of America Picture coding and decoding method and system
JP3807342B2 (en) 2002-04-25 2006-08-09 三菱電機株式会社 Digital signal encoding apparatus, digital signal decoding apparatus, digital signal arithmetic encoding method, and digital signal arithmetic decoding method
JP2003319391A (en) 2002-04-26 2003-11-07 Sony Corp Encoding apparatus and method, decoding apparatus and method, recording medium, and program
PT1467491E (en) 2002-05-02 2007-03-30 Fraunhofer Ges Forschung Arithmetical coding of transform coefficients
US20040234143A1 (en) 2002-07-02 2004-11-25 Makoto Hagai Image encoding method and picture decoding method
US6795584B2 (en) 2002-10-03 2004-09-21 Nokia Corporation Context-based adaptive variable length coding for adaptive block transforms
JP2004297768A (en) 2003-03-10 2004-10-21 Mitsubishi Electric Corp Video signal encoding apparatus, and video signal encoding method
EP1603338A4 (en) 2003-03-10 2007-12-05 Mitsubishi Electric Corp Video signal encoding device and video signal encoding method
US7945102B2 (en) 2003-03-24 2011-05-17 Sony Corporation Data encoding apparatus, data encoding method, data output apparatus, data output method, signal processing system, signal processing apparatus, signal processing method, data decoding apparatus, and data decoding method
US7630440B2 (en) 2003-05-28 2009-12-08 Broadcom Corporation Context adaptive binary arithmetic code decoding engine
US7426308B2 (en) 2003-07-18 2008-09-16 Microsoft Corporation Intraframe and interframe interlace coding and decoding
US7724827B2 (en) 2003-09-07 2010-05-25 Microsoft Corporation Multi-layer run level encoding and decoding
US7379608B2 (en) 2003-12-04 2008-05-27 Fraunhofer-Gesellschaft Zur Foerderung Der Angewandten Forschung, E.V. Arithmetic coding for transforming video and picture data units
TWI330976B (en) 2003-12-05 2010-09-21 Trident Microsystems Far East Method and apparatus for encoding/decoding dynamic graphic content
JP2005184042A (en) 2003-12-15 2005-07-07 Sony Corp Image decoding apparatus, image decoding method, and image decoding program
EP1551185A1 (en) 2004-01-05 2005-07-06 Thomson Licensing S.A. Encoding method, decoding method, and encoding apparatus for a digital picture sequence
US20090102973A1 (en) 2004-01-09 2009-04-23 Harris Scott C Video split device
KR100608050B1 (en) 2004-01-14 2006-08-02 삼성전자주식회사 Storage medium storing interactive graphics stream activated by user's demand, reproducing apparatus, and reproducing method therefor
US8175444B2 (en) 2004-01-14 2012-05-08 Samsung Electronics Co., Ltd. Method of reproducing from storage medium storing interactive graphics stream activated in response to user's command
US7599435B2 (en) 2004-01-30 2009-10-06 Fraunhofer-Gesellschaft Zur Foerderung Der Angewandten Forschung E.V. Video frame encoding and decoding
CN1910922B (en) 2004-01-30 2013-04-17 松下电器产业株式会社 Moving picture coding method and moving picture decoding method
US7469070B2 (en) 2004-02-09 2008-12-23 Lsi Corporation Method for selection of contexts for arithmetic coding of reference picture and motion vector residual bitstream syntax elements
JP4273996B2 (en) 2004-02-23 2009-06-03 ソニー株式会社 Image encoding apparatus and method, and image decoding apparatus and method
ES2383655T3 (en) 2004-04-28 2012-06-25 Panasonic Corporation Apparatus for generating moving image flows, encoding apparatus for moving images, multiplexing apparatus for moving images and decoding apparatus for moving images
EP2309754A3 (en) 2004-06-25 2012-02-01 Panasonic Corporation Image encoding method and image decoding method
CN100584023C (en) 2004-07-14 2010-01-20 新加坡科技研究局 Method and equipment for context-based signal coding and decoding
KR20060016243A (en) 2004-08-17 2006-02-22 엘지전자 주식회사 Apparatus and method for displaying multi channel video in monitoring digital video recorder
JP4375305B2 (en) 2004-10-26 2009-12-02 ソニー株式会社 Information processing apparatus, information processing method, recording medium, and program
US20060120461A1 (en) 2004-12-06 2006-06-08 Roy Knight Two processor architecture supporting decoupling of outer loop and inner loop in video decoder
CA2590705A1 (en) 2005-01-14 2006-07-20 Sungkyunkwan University Methods of and apparatuses for adaptive entropy encoding and adaptive entropy decoding for scalable video encoding
WO2006106935A1 (en) 2005-04-01 2006-10-12 Matsushita Electric Industrial Co., Ltd. Image decoding apparatus and image decoding method
EP1911290A4 (en) 2005-07-08 2010-04-28 Lg Electronics Inc Method for modeling coding information of video signal for compressing/decompressing coding information
KR100878827B1 (en) 2005-07-08 2009-01-14 엘지전자 주식회사 Method for modeling coding information of a video signal to compress/decompress the information
RU2371881C1 (en) 2005-07-08 2009-10-27 ЭлДжи ЭЛЕКТРОНИКС ИНК. Method of modelling video signal coding information for information compression/decompression
US20090034857A1 (en) 2005-07-22 2009-02-05 Mitsubishi Electric Corporation Image encoder and image decoder, image encoding method and image decoding method, image encoding program and image decoding program, and computer readable recording medium recorded with image encoding program and computer readable recording medium recorded with image decoding program
US8488889B2 (en) 2005-07-22 2013-07-16 Mitsubishi Electric Corporation Image encoder and image decoder, image encoding method and image decoding method, image encoding program and image decoding program, and computer readable recording medium recorded with image encoding program and computer readable recording medium recorded with image decoding program
US20080130990A1 (en) 2005-07-22 2008-06-05 Mitsubishi Electric Corporation Image encoder and image decoder, image encoding method and image decoding method, image encoding program and image decoding program, and computer readable recording medium recorded with image encoding program and computer readable recording medium recorded with image decoding program
US20080165849A1 (en) 2005-07-22 2008-07-10 Mitsubishi Electric Corporation Image encoder and image decoder, image encoding method and image decoding method, image encoding program and image decoding program, and computer readable recording medium recorded with image encoding program and computer readable recording medium recorded with image decoding program
US8509551B2 (en) 2005-07-22 2013-08-13 Mitsubishi Electric Corporation Image encoder and image decoder, image encoding method and image decoding method, image encoding program and image decoding program, and computer readable recording medium recording with image encoding program and computer readable recording medium recorded with image decoding program
US20090123066A1 (en) 2005-07-22 2009-05-14 Mitsubishi Electric Corporation Image encoding device, image decoding device, image encoding method, image decoding method, image encoding program, image decoding program, computer readable recording medium having image encoding program recorded therein,
US20080130988A1 (en) 2005-07-22 2008-06-05 Mitsubishi Electric Corporation Image encoder and image decoder, image encoding method and image decoding method, image encoding program and image decoding program, and computer readable recording medium recorded with image encoding program and computer readable recording medium recorded with image decoding program
US20080123977A1 (en) 2005-07-22 2008-05-29 Mitsubishi Electric Corporation Image encoder and image decoder, image encoding method and image decoding method, image encoding program and image decoding program, and computer readable recording medium recorded with image encoding program and computer readable recording medium recorded with image decoding program
US20080130989A1 (en) 2005-07-22 2008-06-05 Mitsubishi Electric Corporation Image encoder and image decoder, image encoding method and image decoding method, image encoding program and image decoding program, and computer readable recording medium recorded with image encoding program and computer readable recording medium recorded with image decoding program
US20080123947A1 (en) 2005-07-22 2008-05-29 Mitsubishi Electric Corporation Image encoding device, image decoding device, image encoding method, image decoding method, image encoding program, image decoding program, computer readable recording medium having image encoding program recorded therein
US20080137744A1 (en) 2005-07-22 2008-06-12 Mitsubishi Electric Corporation Image encoder and image decoder, image encoding method and image decoding method, image encoding program and image decoding program, and computer readable recording medium recorded with image encoding program and computer readable recording medium recorded with image decoding program
CN100584025C (en) 2005-08-04 2010-01-20 华为技术有限公司 Arithmetic decoding system and device based on contents self-adaption
US7894523B2 (en) 2005-09-05 2011-02-22 Lg Electronics Inc. Method for modeling coding information of a video signal for compressing/decompressing coding information
US7595743B1 (en) 2005-10-26 2009-09-29 Lsi Corporation System and method for reducing storage requirements for content adaptive binary arithmetic coding
JP5089878B2 (en) 2005-10-28 2012-12-05 パナソニック株式会社 Image encoding device
US9479794B2 (en) 2005-11-10 2016-10-25 Freescale Semiconductor, Inc. Resource efficient video processing via prediction error computational adjustments
CN100440979C (en) 2005-12-15 2008-12-03 清华大学 Self-adaptive two-dimensional variable length decoding method and device based on context
GB0600141D0 (en) * 2006-01-05 2006-02-15 British Broadcasting Corp Scalable coding of video signals
US7983343B2 (en) * 2006-01-12 2011-07-19 Lsi Corporation Context adaptive binary arithmetic decoding for high definition video
US8446960B2 (en) 2006-01-13 2013-05-21 Fraunhofer-Gesellschaft Zur Foerderung Der Angewandten Forschung E.V. Picture coding using adaptive color space transformation
US20070200949A1 (en) 2006-02-21 2007-08-30 Qualcomm Incorporated Rapid tuning in multimedia applications
KR100750165B1 (en) 2006-02-22 2007-08-17 삼성전자주식회사 Cabac encoding method and apparatus using improved context model selection with better compression ratio, and cabac decoding method and apparatus
JP2007300517A (en) 2006-05-02 2007-11-15 Sony Corp Moving image processing method, program of moving image processing method, recording medium with program of moving image processing method recorded thereon and moving image processor
US7656326B2 (en) 2006-06-08 2010-02-02 Via Technologies, Inc. Decoding of context adaptive binary arithmetic codes in computational core of programmable graphics processing unit
US7262722B1 (en) 2006-06-26 2007-08-28 Intel Corporation Hardware-based CABAC decoder with parallel binary arithmetic decoding
JP4787100B2 (en) 2006-07-27 2011-10-05 パナソニック株式会社 Image encoding device
US9883202B2 (en) 2006-10-06 2018-01-30 Nxp Usa, Inc. Scaling video processing complexity based on power savings factor
US8345767B2 (en) * 2006-12-14 2013-01-01 Nec Corporation Video encoding method, video encoding device, and video encoding program
KR20150039215A (en) 2007-01-11 2015-04-09 톰슨 라이센싱 Methods and apparatus for using syntax for the coded_block_flag syntax element and the coded_block_pattern syntax element for the cavlc 4:4:4 intra, high 4:4:4 intra, and high 4:4:4 predictive profiles in mpeg-4 avc high level coding
US8204129B2 (en) 2007-03-27 2012-06-19 Freescale Semiconductor, Inc. Simplified deblock filtering for reduced memory access and computational complexity
US8750377B2 (en) 2007-04-12 2014-06-10 Thomson Licensing Method and apparatus for context dependent merging for skip-direct modes for video encoding and decoding
JP2008289125A (en) 2007-04-20 2008-11-27 Panasonic Corp Arithmetic decoding apparatus and method thereof
CN100531386C (en) 2007-06-15 2009-08-19 上海富瀚微电子有限公司 Self-adaptive context binary arithmetic encoder and encoding method
RU2339181C1 (en) * 2007-06-25 2008-11-20 Государственное образовательное учреждение высшего профессионального образования Академия Федеральной службы охраны Российской Федерации (Академия ФСО России) Method of additional information transmission at fractal coding of images
US7991237B2 (en) 2007-06-28 2011-08-02 Mitsubishi Electric Corporation Image encoding device, image decoding device, image encoding method and image decoding method
JP5658033B2 (en) 2007-09-02 2015-01-21 エルジー エレクトロニクス インコーポレイティド Video signal processing method and apparatus
TWI339074B (en) 2007-09-05 2011-03-11 Via Tech Inc Operating method and device of coded block flag parameter
US7777654B2 (en) 2007-10-16 2010-08-17 Industrial Technology Research Institute System and method for context-based adaptive binary arithematic encoding and decoding
CN101478672B (en) 2008-01-04 2012-12-19 华为技术有限公司 Video encoding, decoding method and apparatus, video processing system
US8542748B2 (en) 2008-03-28 2013-09-24 Sharp Laboratories Of America, Inc. Methods and systems for parallel video encoding and decoding
JP4962400B2 (en) 2008-04-30 2012-06-27 ソニー株式会社 Arithmetic decoding device
KR20090129926A (en) 2008-06-13 2009-12-17 삼성전자주식회사 Method and apparatus for image encoding by dynamic unit grouping, and method and apparatus for image decoding by dynamic unit grouping
FR2932911A1 (en) 2008-06-24 2009-12-25 France Telecom METHOD AND DEVICE FOR FILLING THE OCCULTATION ZONES OF A DEPTH CARD OR DISPARITIES ESTIMATED FROM AT LEAST TWO IMAGES.
MX2011000261A (en) * 2008-07-10 2011-02-22 Mitsubishi Electric Corp Image encoding device, image decoding device, image encoding method, and image decoding method.
JP5373910B2 (en) 2008-08-19 2013-12-18 トムソン ライセンシング CABAC stream compliance
US8634457B2 (en) 2008-09-26 2014-01-21 Qualcomm Incorporated Determining availability of video data units
US7932843B2 (en) * 2008-10-17 2011-04-26 Texas Instruments Incorporated Parallel CABAC decoding for video decompression
EP2182732A1 (en) * 2008-10-28 2010-05-05 Panasonic Corporation Switching between scans in image coding
WO2010125606A1 (en) 2009-04-29 2010-11-04 Aspa-Japan Co., Ltd. Method for image data compression and compression system of image data
US9100649B2 (en) 2010-02-10 2015-08-04 Lg Electronics Inc. Method and apparatus for processing a video signal
US9237355B2 (en) 2010-02-19 2016-01-12 Qualcomm Incorporated Adaptive motion resolution for video coding
CN102823248B (en) 2010-04-08 2015-06-24 株式会社东芝 Image encoding method and image decoding method
CN102215396A (en) 2010-04-09 2011-10-12 华为技术有限公司 Video coding and decoding methods and systems
US8942282B2 (en) 2010-04-12 2015-01-27 Qualcomm Incorporated Variable length coding of coded block pattern (CBP) in video compression
US20110249754A1 (en) 2010-04-12 2011-10-13 Qualcomm Incorporated Variable length coding of coded block pattern (cbp) in video compression
WO2011128269A1 (en) 2010-04-13 2011-10-20 Fraunhofer-Gesellschaft zur Förderung der angewandten Forschung e.V. Video coding using multi-tree sub - divisions of images
EP2559240B1 (en) 2010-04-13 2019-07-10 GE Video Compression, LLC Inter-plane prediction
PL3429205T3 (en) 2010-05-04 2021-02-08 Lg Electronics Inc. Method and apparatus for processing a video signal
US20110310976A1 (en) 2010-06-17 2011-12-22 Qualcomm Incorporated Joint Coding of Partition Information in Video Coding
DK2897365T3 (en) 2010-07-09 2017-02-13 Samsung Electronics Co Ltd Device for decoding video using block merging
KR20120016980A (en) 2010-08-17 2012-02-27 한국전자통신연구원 Image encoding and decoding method and the same apparatus
CN108737843B (en) 2010-09-27 2022-12-27 Lg 电子株式会社 Method for dividing block and decoding device
US8885704B2 (en) 2010-10-01 2014-11-11 Qualcomm Incorporated Coding prediction modes in video coding
EP3703373B1 (en) 2010-10-08 2024-04-17 GE Video Compression, LLC Picture coding supporting block partitioning and block merging
SI2636218T1 (en) 2010-11-04 2021-12-31 Ge Video Compression, Llc Picture coding supporting block merging and skip mode
US9008181B2 (en) 2011-01-24 2015-04-14 Qualcomm Incorporated Single reference picture list utilization for interprediction video coding
CN107197300B (en) 2011-02-09 2020-03-06 Lg 电子株式会社 Method of encoding and decoding image and apparatus using the same
US20120224639A1 (en) 2011-03-03 2012-09-06 General Instrument Corporation Method for interpolating half pixels and quarter pixels
US9788019B2 (en) 2011-03-09 2017-10-10 Hfi Innovation Inc. Method and apparatus of transform unit partition with reduced complexity
ES2944310T3 (en) 2011-03-21 2023-06-20 Lg Electronics Inc Selecting a motion vector predictor
CN107277539B (en) 2011-05-27 2020-11-06 寰发股份有限公司 Method and apparatus for reducing line cache in context adaptive entropy decoding
US8995523B2 (en) 2011-06-03 2015-03-31 Qualcomm Incorporated Memory efficient context modeling
WO2012167711A1 (en) 2011-06-10 2012-12-13 Mediatek Inc. Method and apparatus of scalable video coding
CN103609113B (en) 2011-06-23 2017-03-08 太阳专利托管公司 Picture decoding method and picture decoding apparatus
CA2830046C (en) 2011-06-24 2018-09-04 Panasonic Corporation Image decoding method, image coding method, image decoding apparatus, image coding apparatus, and image coding and decoding apparatus
EP2725792B1 (en) 2011-06-24 2018-08-22 Sun Patent Trust Image decoding method and image decoding device
CA2842646C (en) 2011-06-27 2018-09-04 Panasonic Corporation Image decoding method, image coding method, image decoding apparatus, image coding apparatus, and image coding and decoding apparatus
RU2602671C2 (en) 2011-06-28 2016-11-20 Сан Пэтент Траст Image decoding method, image encoding method, image decoding device, image encoding device, and image encoding and decoding device
KR102006032B1 (en) 2011-06-29 2019-07-31 선 페이턴트 트러스트 Image decoding method, image encoding method, image decoding device, image encoding device, and image encoding/decoding device
AU2012277220B2 (en) * 2011-06-30 2016-08-11 Sun Patent Trust Image decoding method, image coding method, image decoding apparatus, image coding apparatus, and image coding and decoding apparatus
WO2013001769A1 (en) 2011-06-30 2013-01-03 パナソニック株式会社 Image decoding method, image encoding method, image decoding device, image encoding device, and image encoding/decoding device
US20130003858A1 (en) 2011-06-30 2013-01-03 Vivienne Sze Simplified Context Selection For Entropy Coding of Transform Coefficient Syntax Elements
WO2013005968A2 (en) 2011-07-01 2013-01-10 삼성전자 주식회사 Method and apparatus for entropy encoding using hierarchical data unit, and method and apparatus for decoding
KR20140034209A (en) 2011-07-11 2014-03-19 파나소닉 주식회사 Image decoding method, image encoding method, image decoding apparatus, image encoding apparatus, and image encoding/decoding apparatus
US8804816B2 (en) 2011-08-30 2014-08-12 Microsoft Corporation Video encoding enhancements
US20130188741A1 (en) 2012-01-21 2013-07-25 General Instrument Corporation Devices and methods for sample adaptive offset coding and/or selection of band offset parameters
DK2854398T3 (en) 2012-06-11 2018-01-08 Samsung Electronics Co Ltd Method of encoding videos that share SAO parameters by color component

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101873492A (en) * 2002-08-08 2010-10-27 松下电器产业株式会社 Image encoding method and device
CN101815224A (en) * 2005-07-22 2010-08-25 三菱电机株式会社 Picture coding device and method and picture decoding apparatus and method
EP1827030A2 (en) * 2006-01-10 2007-08-29 Samsung Electronics Co., Ltd. Method and apparatus for changing codec to reproduce video and/or audio data streams encoded by different codecs within a channel
CN101600104A (en) * 2008-06-02 2009-12-09 联发科技股份有限公司 The decoding unit of context adaptive binary arithmetic coding and coding/decoding method

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
THOMAS WIEGAND ET AL: "WD3: Working Draft 3 of High-Efficiency Video Coding (JCTVC-E603)", 《JOINT COLLABORATIVE TEAM ON VIDEO CODING (JCT-VC) OF ITU-T SG16 WP3 AND ISO/IEC JTC1/SC29/WG11》 *

Cited By (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US11930185B2 (en) 2018-11-06 2024-03-12 Beijing Bytedance Network Technology Co., Ltd. Multi-parameters based intra prediction
CN113170122A (en) * 2018-12-01 2021-07-23 北京字节跳动网络技术有限公司 Parameter derivation for intra prediction
US11902507B2 (en) 2018-12-01 2024-02-13 Beijing Bytedance Network Technology Co., Ltd Parameter derivation for intra prediction
US11595687B2 (en) 2018-12-07 2023-02-28 Beijing Bytedance Network Technology Co., Ltd. Context-based intra prediction
US11729405B2 (en) 2019-02-24 2023-08-15 Beijing Bytedance Network Technology Co., Ltd. Parameter derivation for intra prediction

Also Published As

Publication number Publication date
BR112013030562B1 (en) 2022-05-03
US20190320176A1 (en) 2019-10-17
TWI564819B (en) 2017-01-01
US11792400B2 (en) 2023-10-17
US20190075295A1 (en) 2019-03-07
PL2728869T3 (en) 2022-02-28
MX339141B (en) 2016-05-13
AU2012277220B2 (en) 2016-08-11
CA2837537C (en) 2019-04-02
CN103583048B (en) 2017-05-17
US20140301444A1 (en) 2014-10-09
US9154780B2 (en) 2015-10-06
MY167090A (en) 2018-08-10
US10165277B2 (en) 2018-12-25
MX2013013209A (en) 2014-02-20
WO2013001770A1 (en) 2013-01-03
US8792739B2 (en) 2014-07-29
US11356666B2 (en) 2022-06-07
US20170064304A1 (en) 2017-03-02
JP6044860B2 (en) 2016-12-14
JP6436367B2 (en) 2018-12-12
US9794571B2 (en) 2017-10-17
US20220264097A1 (en) 2022-08-18
JP6631893B2 (en) 2020-01-15
US20130004092A1 (en) 2013-01-03
JP5905884B2 (en) 2016-04-20
BR112013030562A8 (en) 2017-07-11
JP2018026869A (en) 2018-02-15
CA2837537A1 (en) 2013-01-03
KR20140031274A (en) 2014-03-12
CN107181948B (en) 2020-03-13
US20170374367A1 (en) 2017-12-28
JP2016158258A (en) 2016-09-01
EP2728869B1 (en) 2021-11-10
EP3944617A1 (en) 2022-01-26
US20200177884A1 (en) 2020-06-04
CN103583048A (en) 2014-02-12
US20230421767A1 (en) 2023-12-28
JP2017085573A (en) 2017-05-18
US10382760B2 (en) 2019-08-13
JP6233760B2 (en) 2017-11-22
RU2597473C2 (en) 2016-09-10
KR102007047B1 (en) 2019-08-02
BR112013030562A2 (en) 2016-09-27
EP2728869A1 (en) 2014-05-07
US10595022B2 (en) 2020-03-17
US20150373335A1 (en) 2015-12-24
JPWO2013001770A1 (en) 2015-02-23
US9525881B2 (en) 2016-12-20
TW201316777A (en) 2013-04-16
EP2728869A4 (en) 2014-12-03
JP2019036990A (en) 2019-03-07
ES2908152T3 (en) 2022-04-27
AU2012277220A1 (en) 2013-12-12
RU2013152614A (en) 2015-09-20

Similar Documents

Publication Publication Date Title
CN103583048B (en) Image decoding method, image encoding method, image decoding device, image encoding device, and image encoding/decoding device
CN103765885B (en) Image decoding method, image encoding method, image decoding apparatus, image encoding apparatus, and image encoding/decoding apparatus
CN103563375B (en) Picture decoding method, method for encoding images, picture decoding apparatus, picture coding device and image encoding/decoding device
CN103563377B (en) Decoding method and decoding device
CN103535036B (en) Coding/decoding method and decoding apparatus
CN103609113B (en) Picture decoding method and picture decoding apparatus
CN103004208B (en) Method for encoding images and device, picture decoding method and device
CN103609110B (en) Picture decoding method, method for encoding images, picture decoding apparatus, picture coding device and image encoding/decoding device
CN103563384B (en) Dynamic image encoding method, dynamic image encoding device, dynamic image decoding method, moving image decoding apparatus and moving image coding and decoding device
CN103563385B (en) Image encoding method, image encoding device, image decoding method, image decoding device, and image encoding/decoding device
CN103718558B (en) Video encoding method and apparatus, video decoding method and apparatus, and video encoding/decoding apparatus
CN103828373B (en) Picture decoding method and picture decoding apparatus
CN103202020B (en) Video decoding method and video decoding apparatus
CN106878732A (en) Picture decoding method and picture decoding apparatus
TWI587686B (en) Decoding method, encoding method, decoding apparatus, coding apparatus, and encoding and decoding apparatus (2)
CN107197284A (en) Image encoding/decoding device
CN107820092A (en) Dynamic image encoding method and device, moving image coding and decoding device

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant